Clinical-radiomics models based on plain X-rays for prediction of lung metastasis in patients with osteosarcoma

OBJECTIVES

Osteosarcoma (OS) is the most common primary malignant bone tumor in adolescents. Lung metastasis (LM) occurs in more than half of patients at different stages of the disease course, which is one of the important factors affecting the long-term survival of OS. To develop and validate machine learning radiomics model based on radiographic and clinical features that could predict LM in OS within 3 years.

METHODS

486 patients (LM = 200, non-LM = 286) with histologically proven OS were retrospectively analyzed and divided into a training set (n = 389) and a validation set (n = 97). Radiographic features and risk factors (sex, age, tumor location, etc.) associated with LM of patients were evaluated. We built eight clinical-radiomics models (k-nearest neighbor [KNN], logistic regression [LR], support vector machine [SVM], random forest [RF], Decision Tree [DT], Gradient Boosting Decision Tree [GBDT], AdaBoost, and extreme gradient boosting [XGBoost]) and compared their performance. The area under the receiver operating characteristic curve (AUC) and accuracy (ACC) were used to evaluate different models.

RESULTS

The radscore, ALP, and tumor size had significant differences between the LM and non-LM groups (t_radscore = -5.829, χ²_ALP = 97.137, t_size = -3.437, P < 0.01). Multivariable LR analyses showed that ALP was an important indicator for predicting LM of OS (odds ratio [OR] = 7.272, P < 0.001). Among the eight models, the SVM-based clinical-radiomics model had the best performance in the validation set (AUC = 0.807, ACC = 0.784).

CONCLUSION

The clinical-radiomics model had good performance in predicting LM in OS, which would be helpful in clinical decision-making.

Risk and Prognostic Factors for Different Organ Metastasis in Primary Osteosarcoma: A Large Population-Based Analysis

OBJECTIVE

Based on a large public cohort, we aimed to investigate the prevalence of distant metastases in patients with osteosarcoma, to evaluate the survival of patients with different metastases and to reveal the related risk and prognostic factors for distant metastases.

METHODS

The information of osteosarcoma patients with or without distant metastases was retrospectively extracted from the Surveillance, Epidemiology, and End Result database from January 2010 to December 2015. Patients were excluded if they were diagnosed at autopsy or via death certification. The Kaplan-Meier method was used to calculate the overall survival in the entire cohort and across patients with metastases to different organs. The related prognostic factors were investigated by univariate and multivariate Cox proportional hazard regression analysis. The logistic regression method was used to reveal the risk factors for the development of different metastases. The effects of different variables on the survival and prevalence of distant metastases were compared using subgroup analysis. Variables with P < 0.05 in the univariate regression analysis were further examined using multivariate regression analysis.

RESULTS

In total, 1470 osteosarcoma patients (mean age 30 ± 22 years) were included, among which 278 patients (18.9%) were initially diagnosed with distant metastasis. The median follow-up duration was 33.0 (30.2-35.8) months. The lung was the most common metastatic site (83.8%), followed by the bone (21.9%), liver (2.9%), and brain (2.2%). A total of 232 patients (83.5%) presented only one distant metastatic site, while the other 46 patients showed two or more metastatic sites. A lower proportion of metastasis was observed in patients aged from 25 to 59 years [odds ratio (OR) = 0.59; 95% confidence interval (CI): 0.37-0.95]. More metastases were noted in patients with T2/T1 (OR = 1.91; 95% CI: 1.28-2.84), T3/T1 (OR = 4.48; 95% CI: 1.78-11.30) and N1/N0 stages (OR = 6.66; 95% CI: 2.68-16.56). The 1-, 3-, and 5-year overall survival rates for metastatic patients were 57.3% (95% CI: 50.8%-63.8%), 25.3% (95% CI: 18.8%-31.9%), and 18.1% (95% CI: 10.2%-26.0%), respectively. Metastatic patients older than 25 years were prone to have poor survival and a relatively better prognosis (hazard ratio = 0.41; 95% CI: 0.25-0.69) was noticed among those who underwent surgery on the primary site. Different metastatic organs have homogeneous and heterogeneous risk and prognostic factors.

CONCLUSION

The high incidence of initial distant metastasis in osteosarcoma and the inconsistent predictive factors should be given more attention in the clinical management of patients with osteosarcoma.

Catecholamine levels and gene expression of their receptors in tissues of adults with osteosarcoma

AIM

The purpose of this work was to identify and measure catecholamines, their metabolites, and the gene expression of catecholamine receptors in osteosarcoma tissue.

MATERIALS AND METHODS

The levels of 3,4-dihydroxyphenylacetic acid, norepinephrine, serotonin, and 5-hydroxyindoleacetic acid in cancer tissue and in adjacent and non-oncological bone tissue were analysed by high-performance liquid chromatography, and the gene expression of catecholamine receptors and of dopamine β-hydroxylase, monoaminoxidase, ki67, and Runx2 in the osteosarcoma tissue, tissue adjacent to the tumour, non-oncological bone, and human brain tissue was analysed by RT-PCR.

RESULTS

We found significantly higher levels of 3,4-dihydroxyphenylacetic acid and norepinephrine in the cancer sample than in adjacent and non-oncological bone. We found that β-adrenergic receptors and dopaminergic receptors, dopamine β-hydroxylase, ki67, Runx2, and serotonergic receptor gene expression were significantly higher in tumour tissue than in adjacent and non-oncological bone.

CONCLUSION

Catecholamines and their receptors could be potential molecular markers for osteosarcoma progression.

Genetic Variant of NFIB is Associated With the Metastasis of Osteosarcoma in Chinese Population

Variant rs7034162 in NFIB was reported to be associated with metastasis of osteosarcoma in European cases with genome-wide significance. Our purpose was to replicate the association of rs7034162 with the metastasis of osteosarcoma in the Chinese population and to further characterize the expression level of NFIB in osteosarcoma tissues. A total of 321 patients were included in this study. Variant rs7034162 was genotyped for each patient using the Taqman genotyping assay. Fifty-two cases of tumor tissues and adjacent normal tissues were collected during surgery. The χ² test was used to investigate the association of rs7034162 with the metastasis of osteosarcoma. The Student t test was used to compare the gene expression between patients with metastasis and those without metastasis. The messenger RNA expression level of NFIB was then compared among different genotypes of rs7034162 with 1-way analysis of variance test. Ninety-three patients were found to have metastasis. Patients with genotype AA had remarkably higher incidence of metastasis than those with genotype TT (34.4% vs 17.1%, P = .002). Patients with metastasis were found to have significantly higher rate of allele A than those without metastasis (53.2% vs 43.9%, P = .03). The messenger RNA expression of NFIB was significantly lower in tumor tissues of patients with metastasis than in those without metastasis (0.00035 ± 0.00017 vs 0.00063 ± 0.0025, P < .001). Compared to patients with genotype TT, those with genotype AA had remarkably decreased expression of NFIB (0.00033 ± 0.0014 vs 0.00067 ± 0.00037, P = .01). Single-nucleotide polymorphism rs7034162 was associated with metastasis of osteosarcoma in the Chinese population possibly via downregulation of NFIB. Further network analyses revealing the related pathways can help elucidate the molecular mechanism of distant metastasis in patients with osteosarcoma.

Lung metastases at the initial diagnosis of high-grade osteosarcoma: prevalence, risk factors and prognostic factors. A large population-based cohort study

BACKGROUND

Osteosarcoma is the most prevalent malignant bone tumor in children and adolescents. Lung metastases are associated with poor prognosis.

OBJECTIVE

The aim here was to explore the prevalence of and risk and prognostic factors for lung metastases in high-grade osteosarcoma patients.

DESIGN AND SETTING

Retrospective cohort study based on the Surveillance, Epidemiology and End Results (SEER) database in the United States.

METHODS

Data on 1,408 high-grade osteosarcoma patients registered in the SEER database between 2010 and 2015 were extracted. From these, all patients with high-grade osteosarcoma and initial lung metastasis were selected for analysis on risk and prognostic factors for lung metastases. Overall survival was estimated.

RESULTS

There were 238 patients (16.90%) with lung metastases at diagnosis. Axial location, tumor size > 10 cm (odds ratio, OR 3.19; 95% confidence interval, CI: 1.58-6.45), higher N stage (OR 4.84; 95% CI: 1.94-12.13) and presence of bone metastases (OR 8.73; 95% CI: 4.37-17.48) or brain metastases (OR 25.63; 95% CI: 1.55-422.86) were significantly associated with lung metastases. Younger age and surgical treatment (hazard ratio, HR 0.46; 95% CI: 0.30-0.71) favored survival. Median survival was prolonged through primary tumor surgery.

CONCLUSIONS

The factors revealed here may guide lung metastasis screening and prophylactic treatment for osteosarcoma patients. A primary tumor in an axial location, greater primary tumor size, higher lymph node stage and presence of bone or brain metastases were significantly correlated with lung metastases. The elderly group (≥ 60 years) showed significant correlation with poor overall survival. For improved survival among high-grade osteosarcoma patients with lung metastases, aggressive surgery on the primary tumor site should be encouraged.

Increased Risk of Lung Metastases in Patients with Giant Cell Bone Tumors: A Systematic Review

Giant cell tumors of the bone are rare, usually benign, tumors consisting of large, multinucleated bone cells. Remarkably, these tumors are characterized by aggressive growth. They tend to recur frequently and, in rare cases, metastasize to the lungs. Previous studies tried to identify risk factors for lung metastasis by giant cell bone tumors. Those studies reported different results due to a small number of patients. Therefore, a particularly high risk associated with this type of bone tumor prompted this systematic review and meta-analysis to identify risk factors for the development of lung metastases. The risk factors for lung metastasis by giant cell bone tumors searched for in this study were gender, age, lung metastasis and recurrence period, follow-up time, primary or recurrent tumor, Campanacci grading, tumor localization, disease course, treatment of primary and recurrent tumors, and pulmonary metastases treated by surgery, radiation, and chemotherapy. This meta-analysis identified the features outlined above by comparing the groups of patients with giant cell bone tumors and lung metastases with the control group consisting of patients without lung metastases. The search for suitable studies revealed 63 publications with a total of 4,295 patients with giant cell bone tumors. Of these, 247 (5.8%; 95% confidence interval (95%CI) 5.1-6.5%) patients had lung metastases. Further, the risk factors for lung metastases were the following: recurrence (p < 0.0001), lung metastasis time (p < 0.0001), Campanacci grade II (p = 0.028) and grade III (p = 0.006), localization in the lower limbs (p = 0.0007), curettage (p = 0.0005), and local irradiation of the primary tumor (p = 0.008). All studies showed a high-risk bias due to the absence of blinding of the participants, personnel, and outcome assessment. Special attention should be paid to tumor recurrence in the long follow-up time, since more advanced giant cell bone tumors, particularly in lower extremities, tend to reoccur and metastasize to the lung. Surgical treatment and local irradiation should be performed thoughtfully, with extended follow-up periods.