Development and internal validation of machine-learning models for predicting survival in patients who underwent surgery for spinal metastases

Study Design

A retrospective study.

Purpose

This study aimed to develop machine-learning algorithms for predicting survival in patients who underwent surgery for spinal metastasis.

Overview of Literature

This study develops machine-learning models to predict postoperative survival in spinal metastasis patients, filling the gaps of traditional prognostic systems. Utilizing data from 389 patients, the study highlights XGBoost and CatBoost algorithms̓ effectiveness for 90, 180, and 365-day survival predictions, with preoperative serum albumin as a key predictor. These models offer a promising approach for enhancing clinical decision-making and personalized patient care.

Methods

A registry of patients who underwent surgery (instrumentation, decompression, or fusion) for spinal metastases between 2004 and 2018 was used. The outcome measure was survival at postoperative days 90, 180, and 365. Preoperative variables were used to develop machine-learning algorithms to predict survival chance in each period. The performance of the algorithms was measured using the area under the receiver operating characteristic curve (AUC).

Results

A total of 389 patients were identified, with 90-, 180-, and 365-day mortality rates of 18%, 41%, and 45% postoperatively, respectively. The XGBoost algorithm showed the best performance for predicting 180-day and 365-day survival (AUCs of 0.744 and 0.693, respectively). The CatBoost algorithm demonstrated the best performance for predicting 90-day survival (AUC of 0.758). Serum albumin had the highest positive correlation with survival after surgery.

Conclusions

These machine-learning algorithms showed promising results in predicting survival in patients who underwent spinal palliative surgery for spinal metastasis, which may assist surgeons in choosing appropriate treatment and increasing awareness of mortality-related factors before surgery.

Safety and feasibility of ultra-long construct navigated minimally invasive spine surgery with adjuvant radiotherapy in extensive spinal metastasis : a comparative analysis

BACKGROUND

Our study compares the outcomes of extensive spinal metastasis patients treated with Ultra-Long Construct Navigated Minimally Invasive Spine Surgery (UNMISS) with Adjuvant Radiotherapy to those receiving only radiotherapy. Spinal metastasis often necessitates interventions like radiotherapy, chemotherapy, or surgery, with an increasing trend towards surgical management. minimally invasive spine surgery has demonstrated advantages over traditional open surgery, with fewer complications and better postoperative outcomes. Radiotherapy continues as a standard for those unsuitable for surgery.

METHODS

This retrospective study included extensive spinal metastasis patients treated between January 2017 and December 2020. We compared patients undergoing UNMISS in conjunction with radiotherapy to patients receiving radiotherapy alone, evaluating demographic data, disease characteristics, and treatment outcomes (VAS, survival) to establish statistical significance.

RESULTS

Twenty-three patients were included in our study. Fourteen patients underwent UNMISS, and nine patients received radiotherapy alone. There was no difference in baseline characteristics of patients. The longest construct in our case series involved T1 to iliac. Both cohorts showed significant improvement in pain scores post-treatment (p = 0.01). However, the UNMISS group demonstrated significantly lower post-treatment VAS scores (p = 0.003), indicating enhanced pain relief. Survival outcomes did not differ significantly between the two groups.

CONCLUSION

The UNMISS should be considered as an alternative treatment in a patient with symptomatic extensive spinal metastasis. The primary goal of this technique is to stabilize the multiple levels of spinal metastasis and decompression of the neural element if needed. This technique is safe and has a better outcome in pain improvement than the patient who received radiotherapy alone.

Development of Machine-Learning Models to Predict Ambulation Outcomes Following Spinal Metastasis Surgery

STUDY DESIGN

Retrospective cohort study.

PURPOSE

This study aimed to develop machine-learning algorithms to predict ambulation outcomes following surgery for spinal metastasis.

OVERVIEW OF LITERATURE

Postoperative ambulation status following spinal metastasis surgery is currently difficult to predict. The improved ability to predict this important postoperative outcome would facilitate management decision-making and help in determining realistic treatment goals.

METHODS

This retrospective study included patients who underwent spinal metastasis at a university-based medical center in Thailand between January 2009 and November 2021. Collected data included preoperative parameters and ambulatory status 90 and 180 days following surgery. Thirteen machine-learning algorithms, namely, artificial neural network, logistic regression, CatBoost classifier, linear discriminant analysis, extreme gradient boosting, extra trees classifier, random forest classifier, gradient boosting classifier, light gradient boosting machine, naïve Bayes, K-neighbor classifier, Ada boost classifier, and decision tree classifier were developed to predict ambulatory status 90 and 180 days following surgery. Model performance was evaluated using the area under the receiver operating characteristic curve (AUC) and F1-score.

RESULTS

In total, 167 patients were enrolled. The number of patients classified as ambulatory 90 and 180 days following surgery was 140 (81.9%) and 137 (82.0%), respectively. The extreme gradient boosting algorithm was found to most accurately predict 180-day ambulatory outcome (AUC, 0.85; F1-score, 0.90), and the decision tree algorithm most accurately predicted 90-day ambulatory outcome (AUC, 0.94; F1-score, 0.88).

CONCLUSIONS

Machine-learning algorithms were effective in predicting ambulatory status following surgery for spinal metastasis. Based on our data, the extreme gradient boosting and decision tree best predicted postoperative ambulatory status 180 and 90 days after spinal metastasis surgery, respectively.

Relationship between gait parameter and spinal sagittal profiles in asymptomatic subjects

BACKGROUND

To study the gait parameters in asymptomatic volunteers and investigate the correlation between the gait and several radiographic sagittal profiles.

METHODS

Asymptomatic volunteers (20-50 years of age) were included and allocated into three subgroups depending on pelvic incidence (low, normal, and high). Standing whole spine radiographs and gait analysis data were obtained. The Pearson Coefficient Correlation was used to determine the relationship between the gait and radiographic profiles.

RESULTS

A total of 55 volunteers (28 male and 27 females) were included. The mean age was 27.35 ± 6.37 years old. The average sacral slope (SS), pelvic tilt (PT), pelvic incidence (PI), and PI-LL mismatch (PI-LL) were 37.78 ± 6.59, 14.51 ± 9.19 degrees, and 52.29 ± 10.87 degrees and - 0.36 ± 11.41, respectively. The mean velocity and stride of all the volunteers were 119.00 ± 30.12 cm/s and 130.25 ± 7.72 cm, correspondingly. The correlation between each of the radiographical and gait parameters was low (ranging from - 0.24 to 0.26).

CONCLUSION

Gait parameters were not differenced significantly between each of the PI subgroups in asymptomatic volunteers. Spinal sagittal parameters also showed a low correlation with gait parameters.

Surgical outcome of scoliosis in patients with Marfan syndrome

PURPOSE

To evaluate clinical and radiographic outcomes after surgical scoliosis correction and posterior instrumented fusion in SMF patients.

METHODS

A single-center medical database was reviewed to identify MF patients who presented with scoliosis from 2000 to 2015. Patients who underwent spinal fusion surgery were included. Demographic, operative and clinical data were reviewed, and the preoperative, postoperative, and latest follow-up radiographic parameters were compared.

RESULTS

Twelve patients were identified (2 males, 10 females) with an average age at surgery of 14.4 ± 2.6 years. Comorbidities were found in 84.6%. Most patients (90.9%) presented with a right thoracic curve. The average preoperative Cobb angle was 75.6 ± 15.5 degrees. Posterior instrumented spinal fusion was performed in all patients (1 hook/pedicular screw and 11 pedicle screws only). The average follow-up period was 6.8 ± 3.1 years. The mean postoperative Cobb angle after surgery and at the final follow-up was 33.4 ± 18.0 degrees and 35.5 ± 18.4 degrees, respectively. There was a statistically significant difference among the preoperative and postoperative Cobb angles (p < 0.001), but no significant difference among the sagittal angles. Two perioperative complications including superficial wound infection and broken rods were observed.

CONCLUSIONS

Posterior scoliosis correction and instrumented spinal fusion resulted in a satisfactory outcome in MF patients. Perioperative complications are not uncommon; however, no neurological complication or spinal decompensation was observed in this study.

LEVEL OF EVIDENCE

IV.

Hidden Neuropathic Pain in Chronic Low Back Pain: Prevalence, Pattern, and Impact on Quality of Life

Objective: A patient with Neuropathic Pain (NP) may suffer from pure NP or may have mixed nociceptive and neuropathic pain. No previous study has investigated NP among Thai patients with Chronic Low Back Pain (CLBP). This study aimed to investigate the prevalence and clinical presentation of NP, and the impact of NP on Quality of Life (QoL) in Thai Chronic Low Back Pain (CLBP) patients.Materials and Methods: Adult patients with CLBP longer than 3 months were included. NP was detected by painDETECT questionnaire, and NP was defined as a score 19. Demographic data, pain characteristics, treatment, Oswestry Disability Index (ODI), and quality of life score (Short Form 36, SF-36) were recorded.Results: 371 CLBP patients were enrolled. The overall prevalence of neuropathic pain was 50.1% (95% CI: 44.9-55.3%). The prevalence of NP in patients with axial low back pain, back pain with pain radiating above the knee, and back pain with pain radiating below the knee was 28.3%, 58.21%, and 59.5%, respectively. Only 48.9% of patients with NP received neuropathic pain medication. Multivariate analysis showed only older age to be associated with NP (OR: 1.017, 95% CI: 1.002-1.033). NP patients had a significantly higher ODI score. There is no difference in most dimension of SF-36 scores, except marginally higher general health and vitality dimension scores.Conclusion: Prevalence of NP in Thai CLBP patients is high. Additionally, it is undertreated and associated with higher disability especially among patients with radiating pain above the knee. Older age is an independent predictor of NP.

Postoperative outcomes of subaxial cervical spine metastasis: Comparison among the anterior, posterior, and combined approaches

•

The combined therapy produced better clinical results and a longer survival.

•

Perioperative complications are significantly higher with combined approach.

•

Anterior and posterior approach have the same neurological improvement rate.

Background

The incidence of subaxial spinal metastases increases due to longer life expectancy resulting from successful modern cancer treatments. The three most utilized approaches for surgical treatment include the anterior, posterior, and combined approaches. However, despite increasing surgical volume, data on the postoperative complication profiles of different operative approaches for this patient population is scarce.

Methods

The institutional databases of two large referral centers in Thailand were retrospectively reviewed. Patients with subaxial cervical spine metastasis who underwent cervical surgery during 2005 to 2015 were identified and enrolled. Clinical presentations, baseline characteristics, operative approach, perioperative complications, and postoperative outcomes, including pain, neurological recovery, and survival, were compared among the three surgical approaches.

Results

The 70 patients (44 with anterior approach, 14 with posterior approach, 12 with combined approach) were enrolled. There were no statistically significant differences in preoperative characteristics, including Charlson Comorbidity Index (CCI), Tomita score, and Revised Tokuhashi score, among the three groups. There were no significant differences among groups for medical complications, surgical complications, neurological recovery, verbal pain score improvement, survival time, or ambulatory status improvement. However, the combined approach did show a significantly higher rate of overall perioperative complications (p = 0.01), intraoperative blood loss (p < 0.001), and operative time (p < 0.001) compared to the other two approaches.

Conclusions

Patients in the combined approach group had the highest rates of perioperative complications. However, although the differences were not statistically significant, patients in the combined group tended to have better clinical outcomes after follow-up and the longest survival time.

Pigmented epithelioid melanocytoma (PEM) of the spine with compression fracture: case report

Background

Pigmented epithelioid melanocytoma (PEM) is a sporadic type of pigmented melanocytic tumor with uncertain malignant potential. PEM arises as a solitary neoplasm that predominantly occurs spontaneously in otherwise healthy patients. Due to its rarity, a gold standard treatment regimen does not exist; however, symptomatic cases should be managed with radiotherapy and surgery.

Case presentation

A 28-year-old Thai female presented with a sudden onset of back pain and weakness of the lower extremities during the postpartum period. Magnetic resonance imaging demonstrated abnormal soft tissue formation from T4 to T7; it extended to the vertebral bodies, left neural foramina, and posterior columns of T6 and T7. The patient underwent complete tumor debulking, decompressive laminectomy from T4 to T8, and posterior instrumentation from T3 to T10. The histopathology and immunohistochemistry suggested PEM. The patient fully resolved back pain after surgery. Nevertheless, as the patient re-presented with a neurological deficit a few months after the operative intervention, it was decided to perform a surgical resection via an en bloc vertebrectomy. At the one-year follow-up, although the patient reported continued improvement of her back pain, there was no motor power improvement.

Conclusions

Spinal cord compression due to PEM is uncommon, especially in adults. Early diagnosis and treatment provide a good prognosis and help to regain lost neurological functions. Complete tumor removal and decompression of the spinal cord must be considered as a treatment strategy. Perioperative radiotherapy and chemotherapy have also been highlighted as treatment modalities for spinal tumors. With our reported case, early operative intervention coupled with radiotherapy produced satisfying outcomes.

Epidemiologic Study of Operative Treatment for Spinal Metastasis in Thailand : A Review of National Healthcare Data from 2005 to 2014

Objective

To study the factors relating to operative treatment for spinal metastasis in Thailand during 2005-2014 and to determine the hospital costs, mortality rate, and incidence of perioperative complication.

Methods

Inpatient reimbursement data from 2005 to 2014 was reviewed from three national healthcare organizations, including the National Health Security Office, the Social Security Office, and the Comptroller General's Department. The search criteria were secondary malignant neoplasm of bone and bone marrow patients (International Classification of Diseases 10th revision, Thai modification codes [ICD 10-TM], C79.5 and C79.8) who underwent spinal surgical treatment (ICD 9th revision, clinical modification procedure with extension codes [ICD 9-CM], 03.0, 03.4, 03.09, and 81.0) during 2005-2014. Epidemiology, comorbidity, and perioperative complication were analyzed.

Results

During the study period, the number of spinal metastasis patients who underwent operative treatment was significantly increased from 0.30 to 0.59 per 100000 (p<0.001). More males (56.14%) underwent surgical treatment for spinal metastasis than females. The most common age group was 45-64 (55.1%). The most common primary tumor sites were the unknown origin, lung, breast, prostate, and hepatocellular/bile duct. Interestingly, the proportion of hepatocellular/bile duct, breast, and lung cancer was significantly increased (p<0.001). The number of patients who had comorbidity or in-hospital complication significantly increased over time (p<0.01); however, the in-hospital mortality rate decreased.

Conclusion

During the last decade, operative treatment for spinal metastasis increased in Thailand. The overall in-hospital complication rate increased; however, the in-hospital mortality rate decreased.

Validity and reliability of the Thai version of the Spine Oncology Study Group Outcomes Questionnaire version 2.0 to assess Quality of Life in Patients with Spinal Metastasis

BACKGROUND CONTEXT

The Spine Oncology Study Group Outcomes Questionnaire version 2.0 (SOSGOQ2.0) is a spinal metastasis (SM)-specific quality of life (QoL) questionnaire that was previously reported to have good reliability and validity. There is currently no Thai version of the SOSGOQ 2.0. (TH-SOSGOQ2.0).

PURPOSE

To assess the psychometric properties of the TH-SOSGOQ 2.0.

STUDY DESIGN/SETTING

Cross-sectional study. Faculty of Medicine Siriraj Hospital, Mahidol University.

PATIENT SAMPLE

Patients who were confirmed diagnosis of metastatic spinal disease, age 18 to 75 years, and having already undergone surgery and/or radiotherapy for the treatment of spinal metastasis.

OUTCOME MEASURES

Validity and reliability of the TH-SOSGOQ 2.0 to assess QoL in Patients with SM.

METHODS

Using the forward-backward translation technique, the SOSGOQ2.0 was translated into Thai language to create the TH-SOSGOQ2.0. SM patients were prospectively enrolled and evaluated for patient QoL using both the TH-SOSGOQ2.0 and the EQ-5D-5L (Thai version) at baseline and 3 months after treatment. Construct validity was assessed using multi-trait scaling analysis, confirmatory factor analysis, and correlation with EQ-5D-5L. Test-retest reliability was assessed in a subgroup of patients who took the TH-SOSGOQ2.0 two times one week apart.

RESULTS

Sixty-eight patients (mean age: 57 years; 30 males, 38 females) were included. The Cronbach's alpha values for the total score, physical function, neurological function, pain, mental health, social function, and post-therapy domains were 0.87, 0.89, 0.91, 0.84, 0.82, 0.75, and 0.85, respectively. Good reliability was demonstrated (interclass correlation coefficient range: 0.70-0.84), except for the social function domain (0.60). Regarding concurrent validity, the TH-SOSGOQ2.0 domains demonstrated moderate to good correlation with the corresponding EQ-5D-5L 9 (Thai version) domains (range: -0.32 to -0.78). Physical function was the most well-correlated domain with the EQ-5D-5L (Thai version) (-0.77).

CONCLUSIONS

TH-SOSGOQ2.0 demonstrated good reliability and validity for assessing QoL in Thai SM patients.

Non-fusion palliative spine surgery without reconstruction is safe and effective in spinal metastasis patients: retrospective study

Considering the shorter life expectancy and poorer prognosis of metastatic epidural spinal cord compression patients, anterior reconstruction and fusion may be unnecessary. This study aimed to investigate the outcomes of palliative surgery for metastatic epidural spinal cord compression with neurological deficit among patients who underwent posterior decompression and instrumentation without fusion or anterior reconstruction. This single-center retrospective review included all patients aged > 18 years with thoracic or lumbar spinal metastasis who were surgically treated for metastatic spinal cord compression without fusion or anterior reconstruction at the Department of Orthopaedic Surgery, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand during July 2015 to December 2017. Data from preoperation to the 1-year follow-up, including demographic and clinical data, Frankel classification, pain scores, complication, revision surgery, health-related quality-of-life scores, and survival data, were collected and analyzed. A total of 30 patients were included. The mean age was 59.83 ± 11.73 years, and 20 (66.7%) patients were female. The mean operative time was 208.17 ± 58.41 min. At least one Frankel grade improvement was reported in 53.33% of patients. The pain visual analog scale, the EuroQOL five-dimension five-level utility score, and the Oswestry Disability Index were all significantly improved at a minimum of 3 months after surgery. No intraoperative mortality or instrument-related complication was reported. The mean survival duration was 11.4 ± 8.97 months. Palliative non-fusion surgery without anterior reconstruction may be considered as a preferable choice for treating spinal metastasis patients with spinal cord compression with neurological deficits.

Reliability and Validity of the Thai Version of the Swiss Spinal Stenosis Questionnaire

The original Swiss Spinal Stenosis (SSS) questionnaire was translated into the Thai version and culturally adapt for use in Thailand. The Thai version—SSS is valid and reliable to use in both clinical and research to address the functional, symptom severity, and satisfaction in Thai lumbar spinal stenosis patients.

Study Design..

Cross-sectional study.

Objective..

To evaluate the reliability and validity of the Thai version of the Swiss Spinal Stenosis (SSS) questionnaire.

Summary of Background Data..

The SSS questionnaire was developed to assess symptom severity, physical functional status, and patient satisfaction in spinal stenosis patients that was previously reported to have good reliability and validity. There is currently no Thai version of the SSS questionnaire.

Methods..

The SSS questionnaire was translated into Thai language to create the Thai version of the Swiss Spinal Stenosis (Thai-SSS) questionnaire. Translation was performed according to international standards using a forward-backward translation protocol. Translation was performed by two expert translators and one physician, and the final version was approved by an expert committee. Thai patients with degenerative lumbar spinal stenosis were enrolled and evaluated using the Thai-SSS questionnaire, the Thai version of the Short Form-36 (SF-36), and a visual analogue scale (VAS) before and after treatment. Reliability and validity of the Thai-SSS were assessed via comparison with the SF-36 and the VAS.

Results..

One hundred seven patients were included. Eighty patients underwent decompression and fusion surgery, and the others underwent decompression alone. Cronbach's alpha of all domains of the Thai-SSS showed excellent internal consistency (0.869–0.921). The intraclass correlation coefficient for test–retest reliability was 0.77 (95% CI: 0.65–0.85). Regarding concurrent validity, the physical function scale of the Thai-SSS was strongly correlated with the physical functioning domain of the SF-36 (r = 0.70). The symptom severity domain of the Thai-SSS was strongly correlated with the bodily pain domain of the SF-36 and the VAS (r = 0.76 and 0.78, respectively).

Conclusion..

The Thai-SSS was found to be a valid and reliable tool for evaluating symptom severity, physical function, and patient satisfaction after treatment in Thai patients with spinal stenosis.

Level of Evidence: 4

Nonfusion Muscle-Sparing Technique to Treat Long-Segment Thoracolumbar Extradural Arachnoid Cyst in a Child: A Case Report and Review of the Literature

BACKGROUND

Spinal extradural arachnoid cyst is an uncommon cause of spinal cord compression. It results from a small dural defect that leads to cerebrospinal fluid (CSF) accumulation in the cyst and subarachnoid space, and the cyst can compress the spinal cord and nerve roots. Surgery is the treatment of choice in symptomatic patients, and most cases have an immediately satisfying outcome after surgery.

CASE DESCRIPTION

A 10-year-old Thai boy presented with progressive weakness in both legs. Magnetic resonance imaging (MRI) demonstrated a giant cystic lesion 23 cm in length from T1-L2 level. To our knowledge, this is the longest arachnoid cyst that has been reported in children. The operation included unilateral paraspinal muscle dissection with spinous process translation techniques, T1-L4 laminectomy, total cyst excision, and dural defect repair at the T4 level. The symptoms were fully resolved within 2 months after surgery. However, thoracic kyphosis was slightly increased, but the patient still has a normal sagittal profile.

CONCLUSIONS

Spinal arachnoid cyst is rare, especially in children. Early diagnosis and treatment are essential for improving the chance of full recovery, and for avoiding complication and irreparable defect. Total cyst removal with dural repair remains the best treatment alternative; however, potential complications, especially spinal malalignment after extensive laminectomy, must be considered. Postoperative MRI showed normal CSF flow and no spinal cord compression. The muscle-sparing technique with multiple-level laminectomy without spinal fusion that we used in this case provided a satisfying outcome.