Search

Background It is unclear whether radiomic phenotypes of brain metastases (BM) are related to radiation therapy prognosis. This study assessed whether a convolutional neural network (CNN)-based radiomics model which learned computer tomography (CT) image features with minimal preprocessing, could predict early response of BM to radiosurgery. Materials and methods Tumor images of 110 BM post stereotactic-radiosurgery (SRS) (within 3 months) were assessed (Response Evaluation Criteria in Solid Tumor, version 1.1) as responders (complete or partial response) or non-responders (stable or progressive disease). Datasets were axial planning CT images containing the tumor center, and the tumor response. Datasets were randomly assigned to training, validation, or evaluation groups repeatedly, to create 50 dataset combinations that were classified into five groups of 10 different dataset combinations with the same evaluation datasets. The CNN learned using training-group images and labels. Validation datasets were used to choose the model that best classified evaluation images as responders or non-responders. Results Of 110 tumors, 57 were classified as responders, and 53 as non-responders. The area under the receiver operating characteristic curve (AUC) of each CNN model for 50 dataset combinations ranged from 0.602 [95% confidence interval (CI)=36.5-83.9%] to 0.826 [95% CI, 64.3-100%]. The AUC of ensemble models, which averaged prediction results of 10 individual models within the same group, ranged from 0.761 (95% CI=55.2-97.1%) to 0.856 (95% CI=68.2-100%). Conclusion A CNN-based ensemble radiomics model accurately predicted SRS responses of unlearned BM images. Thus, CNN models are able to predict SRS prognoses from small datasets.

Dedifferentiated chordoma (DC) is defined as a chordoma containing sarcoma components. DC is distinguished from conventional chordoma by the rapidity of tumor growth and the potential for distant metastasis. We report two cases of DC, which are developed in the sacrum. We reviewed the medical records and imaging studies of 2 patients diagnosed with DC and the literature published. In the first case, percutaneous biopsy revealed that it was conventional chordoma in the sacrum. Patient underwent radiation therapy (RT). Six years after the RT, the tumor recurred. Surgical removal was performed and the recurrent tumor was diagnosed as DC in histopathologic examination. In the second case, a patient underwent gross total resection of sacral tumor, which was diagnosed with conventional chordoma. Aggravated tumor was detected after 4 months, and patient underwent reoperation. The second operation revealed the transformation of the tumor into DC. The survival time of the patients after the diagnosis was 10 and 31 months. Dedifferentiated chordoma is a rare and highly aggressive tumor. De novo type exists, but it usually transformed from recurrent chordoma after surgical resection or radiation.

OBJECTIVE The treatment of giant cell tumor (GCT) is mainly performed surgically. However, GCT in spine seems difficult to treat because of the limited surgical accessibility and proximity. In this report, we analyzed the outcome of GCT treatment in spine. METHODS Between 2000 and 2012, 19 patients received treatment for GCT in spine. Median age at their first diagnosis was 31 years, 10 patients were male, and 9 female. Fourteen tumors were located in the sacrum, 1 in cervical, 1 in thoracic and 3 in lumbar spine. As primary treatment, gross total removal (GTR) was done in 6 patients, and subtotal removal (STR) in 13 patients. Radiation therapy (RT) as an adjuvant therapy was performed in 2 cases in GTR group and 10 cases in STR group. RESULTS During the follow-up, 7 patients had local recurrence (36.8%). The average period until recurrence after primary treatment was 14 months. No recurrence was detected in GTR group. Recurrence was noted in 7 out of 13 patients who underwent STR. These differences were statistically significant (p=0.024). A median of recurrence free period (RFP) was 84 months. Also average RFP of the RT group was 112 months, and non-RT group was 65 months. These differences were statistically significant (p=0.041). CONCLUSION Treatment of choice for GCT in spine is a complete removal of tumor without neurological deficits. In case of incomplete removal, radiation therapy may be a useful adjuvant treatment modality.

This study evaluated clinical outcome and safety of radiosurgery using the Cyberknife for treatment of benign spinal tumors. The authors treated 30 benign spinal tumors in 20 patients with the Cyberknife (Accuray, Inc., Sunnyvale, CA, USA) from 2002 to 2008. Among these there were 20 neurogenic tumors, eight hemangioblastomas, and two meningiomas. Four patients with neurofibromatosis (NF) type 2 and four patients with Von Hippel Lindau disease were also included. Radiosurgery was done as primary treatment for 22 lesions, for postoperative residual tumor control for four lesions, and for the remaining four lesions with image-based progression after initial subtotal resection. The distribution of lesions was cervical (18 tumors), thoracic (six), and cauda equina level (six). Follow-up data included imaging studies, clinical findings, and radiotherapy data. Tumor volume ranged from 0.04 to 33.65 cm³ (mean, 4.52 cm³). A 14-33 Gy marginal dose was delivered in 1-5 fractions. The mean follow-up period was 35.6 months (range, 12-84 months). On follow-up, most lesions decreased in size (57%) or remained unchanged (33%). Two lesions initially decreased, then increased later. One lesion increased without response. With regard to clinical aspects, radicular pain and myelopathic pain improved after radiosurgery in most cases (94%). Motor weakness recovered in two out of five patients and recovery of sensory change occurred in four out of ten patients. In two patients, symptoms were aggravated by tumor enlargement and the occurrence of new lesion. Mean spinal cord volumes receiving more than 10 and 8 Gy were 0.40 ± 0.4 and 0.81 ± 0.7 cm³, respectively. Stereotactic radiosurgery (SRS) using the Cyberknife showed the ability to control benign spinal tumors without complication in most cases.

Introduction: Chordoma and chondrosarcoma are rare tumors of the base of the skull and are difficult candidates for surgical treatment. They are also usually resistant to conventional radiation therapy. We report preliminary results of hypofractionated stereotactic radia- tion therapy (SRT) using the Cyberknife system (Accuray Inc., Sunnyvale, Calif., USA) for primary and recurrent chordomas and chondrosarcomas of the skull base and upper cervical region. Material and Methods: Nine pathologically proven chordoma/chondrosarcoma patients underwent Cyberknife treatment, and in 4 patients Cyberknife was performed as a primary adjuvant treatment after operation. Remainder of the patients had previously received conventional radiotherapy except 1 who had received Gamma Knife treatment. The prescribed tumor dose ranged from 21 to 43.6 Gy in three to five fractions. The dosimetric characteristics were evaluated for conformity and coverage indices. Dose volume histograms of both the tumor and the critical structure were obtained, and the dose delivered to a specific volume (25, 50 and 100%) of the critical structure was calculated in each case. Mass response was measured on follow-up MRI scans. Total tumor doses of different fractionation numbers were converted into single session equivalent doses and linear quadratic equivalent doses of conventional radiation for comparison among patient groups. Result: Nosignificant complications were observed during the treatment and early follow-up periods except one instance of transient esophagitis and one instance of otitis. All treatment plans met the criteria for standard protocol of radiosurgery suggested by the Radiation Therapy Oncology Group, specifically in terms of conformity index, which ranged from 1.01 to 1.83. Three plans had a coverage index that was rated as a minor acceptable deviation. All patients were followed from 11 to 30 (median 24) months following the treatment with regular magnetic resonance images, and 4 patients showed mass reduction. Disease progression was not noted in any patient during the above follow-up period except 1 patient who showed asymptomatic recurrence on 27-month follow-up MRI. Dose volume histograms revealed that the relative dose to volume percent of critical structure, measured at 25, 50 and 100%, was apparently lower in the pretreatment surgical decompression group than in the nondecompression group. Two patients developed radiation-induced myelopathy. The delivered radiation dose to the critical structure calculated using the linear-quadratic formula was within the acceptable range in one case and exceeded 70 Gy at 50% volume of the spinal cord in the other. Conclusion: The hypofractionated Cyberknife SRT is effective in generating therapeutic response in these radioresistant tumors, with minimal toxicity during the procedure and early follow-up period. Repeated radiosurgical treatment may also be feasible for tumor recurrences but deserves great caution with respect to the biological effects of the accumulated dose on the adjacent critical structures. Cyberknife SRT may be a potentially valuable treatment option once the long-term results and appropriate dose calculators are optimally defined.

Introduction Perioperative irradiation is often combined with spine tumor surgery. Radiation is known to be detrimental to healing process of bone fusion. We tried to investigate bone fusion rate in spine tumor surgery cases with perioperative radiation therapy (RT) and to analyze significant factors affecting successful bone fusion. Methods Study cohort was 33 patients who underwent spinal tumor resection and bone graft surgery combined with perioperative RT. Their medical records and radiological data were analyzed retrospectively. The analyzed factors were surgical approach, location of bone graft (anterior vs. posterior), kind of graft (autologous graft vs. allograft), timing of RT (preoperative vs. postoperative), interval of RT from operation in cases of postoperative RT (within 1 month vs. after 1 month) radiation dose (above 38 Gy vs. below 38 Gy) and type of radiation therapy (conventional RT vs. stereotactic radiosurgery). The bone fusion was determined on computed tomography images. Result Bone fusion was identified in 19 cases (57%). The only significant factors to affect bony fusion was the kind of graft (75% in autograft vs. 41 in allograft, p=0.049). Other factors proved to be insignificant relating to postoperative bone fusion. Regarding time interval of RT and operation in cases of postoperative RT, the time interval was not significant (p=0.101). Conclusion Spinal fusion surgery which was combined with perioperative RT showed relatively low bone fusion rate (57%). For successful bone fusion, the selection of bone graft was the most important.

We retrospectively analyzed survival, local control rate, and incidence of radiation toxicities after radiosurgery for recurrent metastatic brain lesions whose initial metastases were treated with whole-brain radiotherapy. Various radiotherapeutical indices were examined to suggest predictors of radiation-related neurological dysfunction.In 46 patients, total 100 of recurrent metastases (mean 2.2, ranged 1-10) were treated by CyberKnife radiosurgery at average dose of 23.1 Gy in 1 to 3 fractions. The median prior radiation dose was 32.7 Gy, the median time since radiation was 5.0 months, and the mean tumor volume was 12.4 cm(3). Side effects were expressed in terms of radiation therapy oncology group (RTOG) neurotoxicity criteria.Mass reduction was observed in 30 patients (65%) on MRI. After the salvage treatment, one-year progression-free survival rate was 57% and median survival was 10 months. Age (60 years) and tumor volume affected survival rate (p=0.03, each). Acute (/=1 month) toxicity was observed in 22% of patients, subacute and chronic (6 months) toxicity occurred in 21%, respectively. Less acute toxicity was observed with small tumors (10 cm(3), p=0.03), and less chronic toxicity occurred at lower cumulative doses (100 Gy, p=0.004). "Radiation toxicity factor" (cumulative dose times tumor volume of1,000 Gyxcm(3)) was a significant predictor of both acute and chronic CNS toxicities.Salvage CyberKnife radiosurgery is effective for recurrent brain metastases in previously irradiated patients, but careful evaluation is advised in patients with large tumors and high cumulative radiation doses to avoid toxicity.