RESUMEN
INTRODUCTION: Wide Surgery is the reference treatment for malignant and aggressive benign primary bone tumors in the spine. When located in the lumbar spine, En-Bloc Spondylectomy (EBS) remains a complex challenge. Moreover, surgery is complicated by the presence of the diaphragm in the thoracolumbar junction and the hinderance of the iliac wings at the lumbosacral levels. Therefore, EBS in the lumbar spine frequently requires combined approaches. The purpose of this study is to describe clinical presentation, tumor characteristics and results of a series of 47 consecutive patients affected by malignant primary bone tumors of the lumbar spine who underwent EBS. MATERIALS AND METHODS: 47 patients were reviewed. Complications were distinguished in early and late whether they occurred before or after 30 days from surgery. Overall survival (OS), disease-free survival (DFS) and local recurrence-free survival (LRFS) were calculated by the Kaplan-Meier product-limit method from surgery until relapse or death. RESULTS: 27 patients presented to observation after a first intralesional approach in a non-specialized center. Chordoma was the most represented histotype. Vertebrectomies were: 23 one-level, 10 two-level, 12 three-level and 2 four-level. Reconstructions were always carried out with screws and rods. The main postoperative complication was blood loss, while hardware failure was the main long-term complication. The 5-year LRFS was 75.5%, the 5-year DFS was 54.3%, and 5-year OS was 63.6%. CONCLUSIONS: The surgical margin obtained during the index surgery was statistically associated with Local Recurrence, DFS and OS, underlining the importance of treating patients in reference centers.
Asunto(s)
Vértebras Lumbares , Neoplasias de la Columna Vertebral , Humanos , Neoplasias de la Columna Vertebral/cirugía , Neoplasias de la Columna Vertebral/complicaciones , Vértebras Lumbares/cirugía , Masculino , Persona de Mediana Edad , Femenino , Adulto , Anciano , Adolescente , Adulto Joven , Complicaciones Posoperatorias/epidemiología , Complicaciones Posoperatorias/etiología , Estudios Retrospectivos , Niño , Resultado del Tratamiento , Cordoma/cirugía , Cordoma/mortalidadRESUMEN
BACKGROUND: Atypical cartilaginous tumour (ACT) and high-grade chondrosarcoma (CS) of long bones are respectively managed with active surveillance or curettage and wide resection. Our aim was to determine diagnostic performance of X-rays radiomics-based machine learning for classification of ACT and high-grade CS of long bones. METHODS: This retrospective, IRB-approved study included 150 patients with surgically treated and histology-proven lesions at two tertiary bone sarcoma centres. At centre 1, the dataset was split into training (n = 71 ACT, n = 24 high-grade CS) and internal test (n = 19 ACT, n = 6 high-grade CS) cohorts, respectively, based on the date of surgery. At centre 2, the dataset constituted the external test cohort (n = 12 ACT, n = 18 high-grade CS). Manual segmentation was performed on frontal view X-rays, using MRI or CT for preliminary identification of lesion margins. After image pre-processing, radiomic features were extracted. Dimensionality reduction included stability, coefficient of variation, and mutual information analyses. In the training cohort, after class balancing, a machine learning classifier (Support Vector Machine) was automatically tuned using nested 10-fold cross-validation. Then, it was tested on both the test cohorts and compared to two musculoskeletal radiologists' performance using McNemar's test. FINDINGS: Five radiomic features (3 morphology, 2 texture) passed dimensionality reduction. After tuning on the training cohort (AUC = 0.75), the classifier had 80%, 83%, 79% and 80%, 89%, 67% accuracy, sensitivity, and specificity in the internal (temporally independent) and external (geographically independent) test cohorts, respectively, with no difference compared to the radiologists (p ≥ 0.617). INTERPRETATION: X-rays radiomics-based machine learning accurately differentiates between ACT and high-grade CS of long bones. FUNDING: AIRC Investigator Grant.
Asunto(s)
Neoplasias Óseas , Condrosarcoma , Humanos , Estudios Retrospectivos , Rayos X , Radiómica , Neoplasias Óseas/diagnóstico por imagen , Neoplasias Óseas/patología , Condrosarcoma/diagnóstico por imagen , Condrosarcoma/patología , Imagen por Resonancia Magnética/métodos , Aprendizaje AutomáticoRESUMEN
BACKGROUND: En bloc resection remains the cornerstone treatment for malignant bone tumors affecting the spine. The thoracic spine poses unique challenges because of the proximity of crucial structures. This study assesses outcomes of patients who underwent en bloc spondylectomy for malignant bone tumors at the thoracic level. METHODS: We retrospectively reviewed 85 cases of primary and secondary bone tumors in the thoracic spine, undergoing en bloc spondylectomy from 1996 to 2016. Evaluation encompassed clinical presentation, tumor characteristics, surgical outcomes, complications, survival, and recurrence. RESULTS: Of 85 patients, 40 presented directly, whereas 45 had undergone previous intralesional surgery. Chondrosarcoma and chordoma comprised the most prevalent primary histologic types; thyroid and kidney carcinomas were the most frequent secondary tumors. Pain was reported in 75 patients at diagnosis. Margins were adequate in 54 cases and intralesional in 31. Immediate postoperative deaths amounted to 4. Major complications included substantial blood loss, neurologic deterioration, and paraplegia. The 5-year local recurrence-free survival was 58.7%, significantly influenced by the surgical margin: patients with wide margins experienced a 5-year local recurrence-free survival of 85.7%, whereas those with marginal and intralesional margins had rates of 56.7% and 45.6%, respectively; overall recurrence was 22.3%, with no notable disparities between previously treated and untreated patients. The 5-year overall survival was 63.2% and 56.2% for primary and secondary tumors, respectively. The overall survival was not significantly influenced by surgical margins. CONCLUSIONS: Managing malignant thoracic bone tumors poses significant challenges. This study underscores the criticality of achieving adequate margins, particularly after previous intralesional approaches.
Asunto(s)
Complicaciones Posoperatorias , Neoplasias de la Columna Vertebral , Vértebras Torácicas , Humanos , Masculino , Femenino , Persona de Mediana Edad , Adulto , Vértebras Torácicas/cirugía , Neoplasias de la Columna Vertebral/cirugía , Estudios Retrospectivos , Anciano , Adulto Joven , Adolescente , Complicaciones Posoperatorias/epidemiología , Complicaciones Posoperatorias/etiología , Resultado del Tratamiento , Recurrencia Local de Neoplasia/cirugía , Niño , Condrosarcoma/cirugía , Anciano de 80 o más Años , Cordoma/cirugíaRESUMEN
Background: Adjuvant radiation therapy following vertebrectomy is a major risk factor for local wound complications such as dehiscence, infection, and skin necrosis. In selected cases, well-vascularized coverage and modification of tension forces on the wound might reduce the risk of postoperative complications and reoperations. We aimed to demonstrate a reduction in general and specific complications in patients undergoing vertebral resection and flap coverage compared with vertebral resection alone. Methods: We retrospectively analyzed and collected data from patients diagnosed with a tumor involving the spine and requiring a total or partial posterior vertebrectomy between January 2012 and October 2022, referred to a single tertiary-level orthopedic and trauma center. We included only patients in whom primary closure of the wound was possible but judged to be under excessive tension. Results: A total of 145 patients underwent partial or total vertebrectomy for oncological reasons at our tertiary-level trauma hospital. Among these, 73 patients were eventually included according to the inclusion and exclusion criteria: 53 in the orthopedic group and 20 in the orthoplastic group. Considering only patients undergoing radiation therapy, the orthoplastic group showed significantly lower rates of overall complications (33% versus 69%) than the orthopedic group. Conclusions: Primary flap coverage, especially in patients receiving RT, reduces the risk of postoperative complications and avoids a second reconstructive operation, consequently reducing patient discomfort, length of hospital stay, and healthcare costs.
RESUMEN
Bizarre parosteal osteochondromatous proliferation (BPOP) is a surface-based bone lesion belonging to the group of benign chondrogenic tumors. The aim of this review is to familiarize the readers with imaging features and differential diagnosis of BPOP, also addressing pathological presentation and treatment options. The peak of incidence of BPOP is in the third and fourth decades of life, although it can occur at any age. Hands are the most common location of BPOP (55%), followed by feet (15%) and long bones (25%). On imaging, BPOP appears as a well-marginated mass of heterotopic mineralization arising from the periosteal aspect of the bone. Typical features of BPOP are contiguity with the underlying bone and lack of cortico-medullary continuity, although cortical interruption and medullary involvement have been rarely reported. Histologically, BPOP is a benign bone surface lesion characterized by osteocartilaginous proliferation with disorganized admixture of cartilage with bizarre features, bone and spindle cells. Differential diagnosis includes both benign-such as florid reactive periostitis, osteochondroma, subungual exostosis, periosteal chondroma and myositis ossificans-and malignant lesions-such as periosteal chondrosarcoma and surface-based osteosarcoma. Treatment consists of surgical resection. Local recurrences are common and treated with re-excision.Critical relevance statement Bizarre parosteal osteochondromatous proliferation is a benign mineralized mass arising from the periosteal aspect of bone cortex. Multi-modality imaging characteristics, pathology features and differential diagnosis are here highlighted to familiarize the readers with this entity and offer optimal patient care.
RESUMEN
PURPOSE: To determine diagnostic performance of MRI radiomics-based machine learning for classification of deep-seated lipoma and atypical lipomatous tumor (ALT) of the extremities. MATERIAL AND METHODS: This retrospective study was performed at three tertiary sarcoma centers and included 150 patients with surgically treated and histology-proven lesions. The training-validation cohort consisted of 114 patients from centers 1 and 2 (n = 64 lipoma, n = 50 ALT). The external test cohort consisted of 36 patients from center 3 (n = 24 lipoma, n = 12 ALT). 3D segmentation was manually performed on T1- and T2-weighted MRI. After extraction and selection of radiomic features, three machine learning classifiers were trained and validated using nested fivefold cross-validation. The best-performing classifier according to previous analysis was evaluated and compared to an experienced musculoskeletal radiologist in the external test cohort. RESULTS: Eight features passed feature selection and were incorporated into the machine learning models. After training and validation (74% ROC-AUC), the best-performing classifier (Random Forest) showed 92% sensitivity and 33% specificity in the external test cohort with no statistical difference compared to the radiologist (p = 0.474). CONCLUSION: MRI radiomics-based machine learning may classify deep-seated lipoma and ALT of the extremities with high sensitivity and negative predictive value, thus potentially serving as a non-invasive screening tool to reduce unnecessary referral to tertiary tumor centers.
Asunto(s)
Lipoma , Liposarcoma , Humanos , Estudios Retrospectivos , Imagen por Resonancia Magnética , Liposarcoma/patología , Lipoma/diagnóstico por imagen , Extremidades , Aprendizaje AutomáticoRESUMEN
Extradural malignant primary spinal tumors are rare and outcome data, especially for younger patients, is limited. In a worldwide (11 centers) study (Predictors of Mortality and Morbidity in the Surgical Management of Primary Tumors of the Spine study; ClinicalTrials.gov Identifier NCT01643174) by the AO Spine Knowledge Forum Tumor, patients surgically treated for primary tumors of the spine between 1992 and 2012, were retrospectively analyzed from a prospective database of their medical history. Medical history, tumor characteristics, diagnostics, treatments, cross-sectional survival, and local recurrences were analyzed. Sixty-eight cases (32 f; 36 m), at an average age of 18.6 ± 4.7 years at the time of diagnosis, were identified (median follow-up 2.9 years). The most common entities were Ewing's sarcoma (42.6%). Of the patients, 28% had undergone previous spine tumor surgery in another center (84% with intralesional margins). Resection was considered "Enneking appropriate" (EA) in 47.8% of the cases. Of the patients, 77.9% underwent chemotherapy and 50% radiotherapy. A local recurrence occurred in 36.4%. Over a third of patients died within a 10-year follow-up period. Kaplan-Meier-analysis demonstrated statistically significant overall survival (p = 0.007) and local recurrence rates (p = 0.042) for tumors treated with EA surgery versus Enneking inappropriate surgery. Aggressive resection of extradural primary malignant spinal tumors combined with adjuvant therapy reveals low local recurrence rates and better outcomes overall in younger patients.
RESUMEN
Extradural primary spinal tumors were retrospectively analyzed from a prospective database of 1495 cases. All subjects with benign primary tumors under the age of 25 years, who were enrolled between 1990 and 2012 (Median FU was 2.4 years), were identified. Patient- and case-related characteristics were collected and statistically analyzed. Results: 161 patients (66f;95m; age 17.0 ± 4.7 years at time of diagnosis) were identified. The most common tumors were osteoblastomas n = 53 (32.9%), osteoid osteomas n = 45 (28.0%), and aneurysmal bone cysts n = 32 (19.9%). The tumor grade, according to the Enneking Classification S1/S2/S3, was 14/73/74 (8.7/45.3/46.0%), respectively. Tumor-related pain was present in 156 (96.9%) patients. Diagnosis was achieved by biopsies in 2/3 of the cases. Spinal fixation was used in >50% of the cases. Resection was Enneking appropriate in n = 100 (62.1%) of cases. Local recurrence occurred in 21 (13.1%) patients. Two patients died within a 10-year follow-up period. Conclusion: This is one of the largest international multicenter cohorts of young patients surgically treated for benign spinal tumors. The heterogenic young patient cohort presented at a mid-term follow-up without a correlation between the grade of aggressiveness in resection and local recurrence rates. Further prospective data are required to identify prognostic factors that determine oncological and functional outcomes for young patients suffering from these rare tumors.
RESUMEN
Objective: The extent of response to neoadjuvant chemotherapy predicts survival in Ewing sarcoma. This study focuses on MRI radiomics of skeletal Ewing sarcoma and aims to investigate feature reproducibility and machine learning prediction of response to neoadjuvant chemotherapy. Materials and methods: This retrospective study included thirty patients with biopsy-proven skeletal Ewing sarcoma, who were treated with neoadjuvant chemotherapy before surgery at two tertiary sarcoma centres. 7 patients were poor responders and 23 were good responders based on pathological assessment of the surgical specimen. On pre-treatment T1-weighted and T2-weighted MRI, 2D and 3D tumour segmentations were manually performed. Features were extracted from original and wavelet-transformed images. Feature reproducibility was assessed through small geometrical transformations of the regions of interest mimicking multiple manual delineations, and intraclass correlation coefficient >0.75 defined feature reproducibility. Feature selection also consisted of collinearity and significance analysis. After class balancing in the training cohort, three machine learning classifiers were trained and tested on unseen data using hold-out cross-validation. Results: 1303 (77%) 3D and 620 (65%) 2D radiomic features were reproducible. 4 3D and 4 2D features passed feature selection. Logistic regression built upon 3D features achieved the best performance with 85% accuracy (AUC=0.9) in predicting response to neoadjuvant chemotherapy. Conclusion: Compared to 2D approach, 3D MRI radiomics of Ewing sarcoma had superior reproducibility and higher accuracy in predicting response to neoadjuvant chemotherapy, particularly when using logistic regression classifier.
RESUMEN
PURPOSE: To evaluate stability and machine learning-based classification performance of radiomic features of spine bone tumors using diffusion- and T2-weighted magnetic resonance imaging (MRI). MATERIAL AND METHODS: This retrospective study included 101 patients with histology-proven spine bone tumor (22 benign; 38 primary malignant; 41 metastatic). All tumor volumes were manually segmented on morphologic T2-weighted sequences. The same region of interest (ROI) was used to perform radiomic analysis on ADC map. A total of 1702 radiomic features was considered. Feature stability was assessed through small geometrical transformations of the ROIs mimicking multiple manual delineations. Intraclass correlation coefficient (ICC) quantified feature stability. Feature selection consisted of stability-based (ICC > 0.75) and significance-based selections (ranking features by decreasing Mann-Whitney p-value). Class balancing was performed to oversample the minority (i.e., benign) class. Selected features were used to train and test a support vector machine (SVM) to discriminate benign from malignant spine tumors using tenfold cross-validation. RESULTS: A total of 76.4% radiomic features were stable. The quality metrics for the SVM were evaluated as a function of the number of selected features. The radiomic model with the best performance and the lowest number of features for classifying tumor types included 8 features. The metrics were 78% sensitivity, 68% specificity, 76% accuracy and AUC 0.78. CONCLUSION: SVM classifiers based on radiomic features extracted from T2- and diffusion-weighted imaging with ADC map are promising for classification of spine bone tumors. Radiomic features of spine bone tumors show good reproducibility rates.
Asunto(s)
Neoplasias Óseas , Aprendizaje Automático , Neoplasias Óseas/diagnóstico por imagen , Humanos , Imagen por Resonancia Magnética , Reproducibilidad de los Resultados , Estudios RetrospectivosRESUMEN
BACKGROUND: Atypical cartilaginous tumour (ACT) and grade II chondrosarcoma (CS2) of long bones are respectively managed with watchful waiting or curettage and wide resection. Preoperatively, imaging diagnosis can be challenging due to interobserver variability and biopsy suffers from sample errors. The aim of this study is to determine diagnostic performance of MRI radiomics-based machine learning in differentiating ACT from CS2 of long bones. METHODS: One-hundred-fifty-eight patients with surgically treated and histology-proven cartilaginous bone tumours were retrospectively included at two tertiary bone tumour centres. The training cohort consisted of 93 MRI scans from centre 1 (n=74 ACT; n=19 CS2). The external test cohort consisted of 65 MRI scans from centre 2 (n=45 ACT; n=20 CS2). Bidimensional segmentation was performed on T1-weighted MRI. Radiomic features were extracted. After dimensionality reduction and class balancing in centre 1, a machine-learning classifier (Extra Trees Classifier) was tuned on the training cohort using 10-fold cross-validation and tested on the external test cohort. In centre 2, its performance was compared with an experienced musculoskeletal oncology radiologist using McNemar's test. FINDINGS: After tuning on the training cohort (AUC=0.88), the machine-learning classifier had 92% accuracy (60/65, AUC=0.94) in identifying the lesions in the external test cohort. Its accuracies in correctly classifying ACT and CS2 were 98% (44/45) and 80% (16/20), respectively. The radiologist had 98% accuracy (64/65) with no difference compared to the classifier (p=0.134). INTERPRETATION: Machine learning showed high accuracy in classifying ACT and CS2 of long bones based on MRI radiomic features. FUNDING: ESSR Young Researchers Grant.
Asunto(s)
Neoplasias Óseas , Condrosarcoma , Neoplasias Óseas/diagnóstico por imagen , Neoplasias Óseas/patología , Condrosarcoma/diagnóstico por imagen , Condrosarcoma/patología , Humanos , Aprendizaje Automático , Imagen por Resonancia Magnética/métodos , Estudios RetrospectivosRESUMEN
OBJECTIVE(S): There is still limited data in the literature concerning the survival of patients with tumors of the thoracic spine. In this study, we analyzed clinical features, perioperative and long-term outcomes in patients who underwent vertebrectomy for cancer. Furthermore, we evaluated the survival and surgical complications. METHODS: We retrospectively reviewed all cases of thoracic spinal tumors treated by the same team between 1998 and 2018. We divided them into three groups according to type of tumor (primary vertebral, primary lung and metastases) and compared outcomes. For each patient, Overall Survival (OS) and Cumulative Incidence of Relapse (CIR) were estimated. Complications and survival were analyzed using a logistic model. RESULTS: Seventy-two patients underwent thoracic spine surgery (40 in group 1, 16 in each group 2 and 3). Thirty patients died at the end of the observation at a mean follow up time of 60 months (41%). The 5-year overall survival was 72% (95% CI: 0.52-0.84), 20% (95% CI: 0.05-0.43) and 27% (95% CI: 0.05-0.56) for each group, respectively. CIR of group 3 was higher (HR 2.57, 95% CI: 1.22-5.45, p = 0.013). The logistic model revealed that age was related to complications (p = 0.04), while surgery for a type 3 tumor was related to mortality (p = 0.02). CONCLUSIONS: Although the cohort size was limited, primary vertebral tumors displayed the best 5-y-OS with an acceptable complications rate. The indication of surgery should be advised by a multidisciplinary team and only for selected cases. Finally, the use of a combined approach does not increase the risk of complications.
RESUMEN
Wide resection is currently considered the mainstay treatment for primary bone tumors. When the tumor is located in anatomically complex segments, 3D-Printed Titanium Custom-Made Prostheses (3DPTCMP) are possible reconstructive solutions. The aim of the present paper is to analyze indications, results and complications of a series of 14 patients who underwent pelvis reconstruction with 3DPTCMP after tumor removal from January 2015 to December 2019. Chondrosarcoma was the main histology; indications were tumors located in the acetabular area without enough residual bone to support a cup with an iliac stem, and tumors located near the sacrum-iliac joint. The margins were wide in 12 cases, and marginal and intralesional in one case each. In three cases, resection also included the sacrum-iliac joint, so a spine stabilization was performed and linked to the pelvic prosthesis; The average MSTS score was 46.3%; the 5-year local recurrence-free survival was 85.7%. Wound dehiscences were the main complication, resolved with multiple debridements; nevertheless, prosthesis removal was necessary in one case. Currently, the 3DPTCMP is an effective resource for reconstruction after resection of tumors located in the pelvis. Further studies are necessary to value long-term results; more strategies are necessary to try to reduce the infection rate and improve osteointegration.
RESUMEN
BACKGROUND: En bloc surgery is the mainstay treatment for primary malignant bone tumors, as well as in the cervical spine. Unfortunately, literature on the topic is limited to case reports and small series. METHODS: We reviewed all patients affected by primary cervical spine bone tumors treated with en bloc surgeries from 1996 to 2016 and identified 30 eligible cases. We evaluated the clinical presentation and tumor characteristics and reported surgical results, complications, recurrence, and survival rates. RESULTS: Only 17 of 30 patients had not been previously treated at presentation. Osteosarcoma and chordoma were the most frequent tumors, and pain was reported in all cases. En bloc spondylectomy, hemispondylectomy, and posterior arch en bloc resection were performed in 16, 12, and 2 patients, respectively. The obtained margin was adequate (wide and marginal) in 60% of cases and intralesional in the remaining cases. Two deaths occurred in the immediate postoperative period. Neurological deterioration, dural tear, and dysphagia were the most frequent complications. The 5-year local recurrence-free survival was 70.4%. The recurrence rate was 38.5% and 11.7% in previously and non-previously treated patients, respectively (χ2: 2.94; P = 0.086). Overall survival at 5 years was 58% and 47% for all series and malignant tumors, respectively. CONCLUSION: Primary cervical spine bone tumors present a difficult approach. Findings suggest that patients treated with en bloc surgery show recurrence and survival rates comparable to the same tumors located in the thoracolumbar spine.
Asunto(s)
Vértebras Cervicales/cirugía , Procedimientos Neuroquirúrgicos/métodos , Neoplasias de la Columna Vertebral/secundario , Adolescente , Adulto , Anciano , Niño , Femenino , Humanos , Masculino , Persona de Mediana Edad , Recurrencia Local de Neoplasia/cirugía , Complicaciones Posoperatorias/cirugía , Resultado del Tratamiento , Adulto JovenRESUMEN
OBJECTIVE: Oncological resection of primary spine tumors is associated with lower recurrence rates. However, even in the most experienced hands, the execution of a meticulously drafted plan sometimes fails. The objectives of this study were to determine how successful surgical teams are at achieving planned surgical margins and how successful surgeons are in intraoperatively assessing tumor margins. The secondary objective was to identify factors associated with successful execution of planned resection. METHODS: The Primary Tumor Research and Outcomes Network (PTRON) is a multicenter international prospective registry for the management of primary tumors of the spine. Using this registry, the authors compared 1) the planned surgical margin and 2) the intraoperative assessment of the margin by the surgeon with the postoperative assessment of the margin by the pathologist. Univariate analysis was used to assess whether factors such as histology, size, location, previous radiotherapy, and revision surgery were associated with successful execution of the planned margins. RESULTS: Three hundred patients were included. The surgical plan was successfully achieved in 224 (74.7%) patients. The surgeon correctly assessed the intraoperative margins, as reported in the final assessment by the pathologist, in 239 (79.7%) patients. On univariate analysis, no factor had a statistically significant influence on successful achievement of planned margins. CONCLUSIONS: In high-volume cancer centers around the world, planned surgical margins can be achieved in approximately 75% of cases. The morbidity of the proposed intervention must be balanced with the expected success rate in order to optimize patient management and surgical decision-making.
Asunto(s)
Márgenes de Escisión , Neoplasias de la Columna Vertebral , Estudios de Factibilidad , Humanos , Recurrencia Local de Neoplasia , Estudios Retrospectivos , Neoplasias de la Columna Vertebral/diagnóstico por imagen , Neoplasias de la Columna Vertebral/cirugía , Columna Vertebral , Resultado del TratamientoRESUMEN
BACKGROUND: Clinical management ranges from surveillance or curettage to wide resection for atypical to higher-grade cartilaginous tumours, respectively. Our aim was to investigate the performance of computed tomography (CT) radiomics-based machine learning for classification of atypical cartilaginous tumours and higher-grade chondrosarcomas of long bones. METHODS: One-hundred-twenty patients with histology-proven lesions were retrospectively included. The training cohort consisted of 84 CT scans from centre 1 (n=55 G1 or atypical cartilaginous tumours; n=29 G2-G4 chondrosarcomas). The external test cohort consisted of the CT component of 36 positron emission tomography-CT scans from centre 2 (n=16 G1 or atypical cartilaginous tumours; n=20 G2-G4 chondrosarcomas). Bidimensional segmentation was performed on preoperative CT. Radiomic features were extracted. After dimensionality reduction and class balancing in centre 1, the performance of a machine-learning classifier (LogitBoost) was assessed on the training cohort using 10-fold cross-validation and on the external test cohort. In centre 2, its performance was compared with preoperative biopsy and an experienced radiologist using McNemar's test. FINDINGS: The classifier had 81% (AUC=0.89) and 75% (AUC=0.78) accuracy in identifying the lesions in the training and external test cohorts, respectively. Specifically, its accuracy in classifying atypical cartilaginous tumours and higher-grade chondrosarcomas was 84% and 78% in the training cohort, and 81% and 70% in the external test cohort, respectively. Preoperative biopsy had 64% (AUC=0.66) accuracy (p=0.29). The radiologist had 81% accuracy (p=0.75). INTERPRETATION: Machine learning showed good accuracy in classifying atypical and higher-grade cartilaginous tumours of long bones based on preoperative CT radiomic features. FUNDING: ESSR Young Researchers Grant.
Asunto(s)
Neoplasias Óseas/diagnóstico por imagen , Condrosarcoma/diagnóstico por imagen , Interpretación de Imagen Radiográfica Asistida por Computador/métodos , Anciano , Área Bajo la Curva , Neoplasias Óseas/patología , Condrosarcoma/patología , Femenino , Humanos , Aprendizaje Automático , Masculino , Persona de Mediana Edad , Clasificación del Tumor , Estudios Retrospectivos , Tomografía Computarizada por Rayos XRESUMEN
Introduction: Axial osteosarcoma and Ewing sarcoma are rare, aggressive neoplasms with a worse prognosis than with tumors involving the extremities because they are more likely to be associated with larger tumor volumes, older age, primary metastases, and a poor histological response to chemotherapy. The 5-year OS rates are reportedly in the range of 18-41% for axial osteosarcoma, and 46-64% for Ewing sarcoma.Area covered: The treatment of axial bone tumors is the same as for extremity bone tumors, and includes chemotherapy, surgery and/or radiotherapy.Expert opinion: Local treatment of axial tumors is particularly difficult due to their proximity to neurological and vascular structures, which often makes extensive and en bloc resections impossible without causing significant morbidity. The incidence of local relapse is consequently high, and this is the main issue in the treatment of these tumors. Radiotherapy is an option in the case of surgical resections with close or positive margins, as well as for inoperable tumors. Delivering high doses of RT to the spinal cord can be dangerous. Given the complexity and rarity of these tumors, it is essential for this subset of patients to be treated at selected reference institutions with specific expertise and multidisciplinary skills.
Asunto(s)
Neoplasias Óseas , Osteosarcoma , Sarcoma de Ewing , Neoplasias Óseas/patología , Humanos , Recurrencia Local de Neoplasia , Osteosarcoma/patología , Osteosarcoma/terapia , Pronóstico , Sarcoma de Ewing/patología , Sarcoma de Ewing/terapiaRESUMEN
PURPOSE: Spinal lesion differential diagnosis remains challenging even in MRI. Radiomics and machine learning (ML) have proven useful even in absence of a standardized data mining pipeline. We aimed to assess ML diagnostic performance in spinal lesion differential diagnosis, employing radiomic data extracted by different software. METHODS: Patients undergoing MRI for a vertebral lesion were retrospectively analyzed (nâ¯=â¯146, 67 males, 79 females; mean age 63⯱â¯16 years, range 8-89 years) and constituted the train (nâ¯=â¯100) and internal test cohorts (nâ¯=â¯46). Part of the latter had additional prior exams which constituted a multi-scanner, external test cohort (nâ¯=â¯35). Lesions were labeled as benign or malignant (2-label classification), and benign, primary malignant or metastases (3-label classification) for classification analyses. Features extracted via 3D Slicer heterogeneityCAD module (hCAD) and PyRadiomics were independently used to compare different combinations of feature selection methods and ML classifiers (nâ¯=â¯19). RESULTS: In total, 90 and 1548 features were extracted by hCAD and PyRadiomics, respectively. The best feature selection method-ML algorithm combination was selected by 10 iterations of 10-fold cross-validation in the training data. For the 2-label classification ML obtained 94% accuracy in the internal test cohort, using hCAD data, and 86% in the external one. For the 3-label classification, PyRadiomics data allowed for 80% and 69% accuracy in the internal and external test sets, respectively. CONCLUSIONS: MRI radiomics combined with ML may be useful in spinal lesion assessment. More robust pre-processing led to better consistency despite scanner and protocol heterogeneity.
Asunto(s)
Neoplasias Óseas , Aprendizaje Automático , Adolescente , Adulto , Anciano , Anciano de 80 o más Años , Neoplasias Óseas/diagnóstico por imagen , Niño , Femenino , Humanos , Imagen por Resonancia Magnética , Masculino , Persona de Mediana Edad , Estudios Retrospectivos , Programas Informáticos , Adulto JovenRESUMEN
Breast cancer patients are at a high risk of complications from bone metastasis. Molecular characterization of bone metastases is essential for the discovery of new therapeutic targets. Here, we investigated the expression and the intracellular distribution of KH RNA binding domain containing, signal transduction associated 1 (KHDRBS1), leptin, leptin receptor (LEPR), and adiponectin in bone metastasis from breast carcinoma and looked for correlations between the data. The expression of these proteins is known in breast carcinoma, but it has not been investigated in bone metastatic tissue to date. Immunohistochemical analysis was carried out on bone metastasis specimens, then semiquantitative evaluation of the results and the Pearson test were performed to determine eventual correlations. KHDRBS1 expression was significantly higher in the nuclei than in the cytosol of metastatic cells; LEPR was prevalently observed in the cytosol and the nuclei; leptin and adiponectin were found in metastatic cells and stromal cells; the strongest positive correlation was between nuclear KHDRBS1 and nuclear LEPR expression. Taken together, our findings support the importance of the leptin/LEPR/KHDRBS1 axis and of adiponectin in the progression of bone metastasis and suggest their potential application in pharmacological interventions.
RESUMEN
PURPOSE: To evaluate the diagnostic performance of machine learning for discrimination between low-grade and high-grade cartilaginous bone tumors based on radiomic parameters extracted from unenhanced magnetic resonance imaging (MRI). METHODS: We retrospectively enrolled 58 patients with histologically-proven low-grade/atypical cartilaginous tumor of the appendicular skeleton (nâ¯=â¯26) or higher-grade chondrosarcoma (nâ¯=â¯32, including 16 appendicular and 16 axial lesions). They were randomly divided into training (nâ¯=â¯42) and test (nâ¯=â¯16) groups for model tuning and testing, respectively. All tumors were manually segmented on T1-weighted and T2-weighted images by drawing bidimensional regions of interest, which were used for first order and texture feature extraction. A Random Forest wrapper was employed for feature selection. The resulting dataset was used to train a locally weighted ensemble classifier (AdaboostM1). Its performance was assessed via 10-fold cross-validation on the training data and then on the previously unseen test set. Thereafter, an experienced musculoskeletal radiologist blinded to histological and radiomic data qualitatively evaluated the cartilaginous tumors in the test group. RESULTS: After feature selection, the dataset was reduced to 4 features extracted from T1-weighted images. AdaboostM1 correctly classified 85.7 % and 75 % of the lesions in the training and test groups, respectively. The corresponding areas under the receiver operating characteristic curve were 0.85 and 0.78. The radiologist correctly graded 81.3 % of the lesions. There was no significant difference in performance between the radiologist and machine learning classifier (Pâ¯=â¯0.453). CONCLUSIONS: Our machine learning approach showed good diagnostic performance for classification of low-to-high grade cartilaginous bone tumors and could prove a valuable aid in preoperative tumor characterization.