Comparison of CT and MRI brain tumor imaging using a canine glioma model

BrainNet: a fusion assisted novel optimal framework of residual blocks and stacked autoencoders for multimodal brain tumor classification

A significant issue in computer-aided diagnosis (CAD) for medical applications is brain tumor classification. Radiologists could reliably detect tumors using machine learning algorithms without extensive surgery. However, a few important challenges arise, such as (i) the selection of the most important deep learning architecture for classification (ii) an expert in the field who can assess the output of deep learning models. These difficulties motivate us to propose an efficient and accurate system based on deep learning and evolutionary optimization for the classification of four types of brain modalities (t1 tumor, t1ce tumor, t2 tumor, and flair tumor) on a large-scale MRI database. Thus, a CNN architecture is modified based on domain knowledge and connected with an evolutionary optimization algorithm to select hyperparameters. In parallel, a Stack Encoder-Decoder network is designed with ten convolutional layers. The features of both models are extracted and optimized using an improved version of Grey Wolf with updated criteria of the Jaya algorithm. The improved version speeds up the learning process and improves the accuracy. Finally, the selected features are fused using a novel parallel pooling approach that is classified using machine learning and neural networks. Two datasets, BraTS2020 and BraTS2021, have been employed for the experimental tasks and obtained an improved average accuracy of 98% and a maximum single-classifier accuracy of 99%. Comparison is also conducted with several classifiers, techniques, and neural nets; the proposed method achieved improved performance.

Application of Deep Learning Technology in Glioma

A common and most basic brain tumor is glioma that is exceptionally dangerous to health of various patients. A glioma segmentation, which is primarily magnetic resonance imaging (MRI) oriented, is considered as one of common tools developed for doctors. These doctors use this system to examine, analyse, and diagnose appearance of the glioma's outward for both patients, i.e., indoor and outdoor. In the literature, a widely utilized approach for the segmentation of glioma is the deep learning-oriented method. To cope with this issue, a segmentation of glioma approach, i.e., primarily on the convolution neural networks, is developed in this manuscript. A DM-DA-enabled cascading approach for the segmentation of glioma, which is 2DResUnet-enabled model, is reported to resolve the problem of spatial data acquisition of insufficient 3D specifically in the 2D full CNN along with the core issue of memory consumption of 3D full CNN. For gliomas segmentation at various stages, we have utilized multiscale fusion approach, attention, segmentation, and DenseBlock. Moreover, for reducing three dimensionalities of the Unet model, a sampling of fixed region is used along with multisequence data of the glioma image. Finally, the CNN model has the ability of producing a better segmentation of tumor preferably with minimum possible memory. The proposed model has used BraTS18 and BraTS17 benchmark data sets for fivefold cross-validation (local) and online evaluation preferably official, respectively. Evaluation results have verified that edema's Dice Score preferable average, enhancement, and core areas of the segmentation of the glioma with DM-DA-Unet perform exceptionally well on the validation set of BraTS17. Finally, average sensitivity was observed to be high as well, which is approximately closer to the best segmentation model and its effect on the validation set of BraTS1 and has segmented gliomas accurately.

Comparison of response assessment in veterinary neuro-oncology and two volumetric neuroimaging methods to assess therapeutic brain tumour responses in veterinary patients

Standardized veterinary neuroimaging response assessment methods for brain tumours are lacking. Consequently, a response assessment in veterinary neuro-oncology (RAVNO) system which uses the sum product of orthogonal lesion diameters on 1-image section with the largest tumour area, has recently been proposed. In this retrospective study, 22 pre-treatment magnetic resonance imaging (MRI) studies from 18 dogs and four cats with suspected intracranial neoplasia were compared by a single observer to 32 post-treatment MRIs using the RAVNO system and two volumetric methods based on tumour margin or area delineation with HOROS and 3D Slicer software, respectively. Intra-observer variability was low, with no statistically significant differences in agreement index between methods (mean AI ± SD, 0.91 ± 0.06 for RAVNO; 0.86 ± 0.08 for HOROS; and 0.91 ± 0.05 for 3D slicer), indicating good reproducibility. Response assessments consisting of complete or partial responses, and stable or progressive disease, agreed in 23 out of 32 (72%) MRI evaluations using the three methods. The RAVNO system failed to identify changes in mass burden detected with volumetric methods in six cases. 3D Slicer differed from the other two methods in three cases involving cysts or necrotic tissue as it allowed for more accurate exclusion of these structures. The volumetric response assessment methods were more precise in determining changes in absolute tumour burden than RAVNO but were more time-consuming to use. Based on observed agreement between methods, low intra-observer variability and decreased time constraint, RAVNO might be a suitable response assessment method for the clinical setting.

Multiparametric magnetic resonance imaging features of a canine glioblastoma model

PURPOSE

To assess glioblastoma multiforme (GBM) formation with similar imaging characteristics to human GBM using multiparametric magnetic resonance imaging (MRI) in an orthotopic xenograft canine GBM model.

MATERIALS AND METHODS

The canine GBM cell line J3T1 was subcutaneously injected into 6-week-old female BALB/c nude mice to obtain tumour fragments. Tumour fragments were implanted into adult male mongrel dog brains through surgery. Multiparametric MRI was performed with conventional MRI, diffusion-weighted imaging, and dynamic susceptibility contrast-enhanced perfusion-weighted imaging at one week and two weeks after surgery in a total of 15 surgical success cases. The presence of tumour cells, the necrotic area fraction, and the microvessel density (MVD) of the tumour on the histologic specimen were assessed. Tumour volume, diffusion, and perfusion parameters were compared at each time point using Wilcoxon signed-rank tests, and the differences between tumour and normal parenchyma were compared using unpaired t-tests. Spearman correlation analysis was performed between the imaging and histologic parameters.

RESULTS

All animals showed a peripheral enhancing lesion on MRI and confirmed the presence of a tumour through histologic analysis (92.3%). The normalized perfusion values did not show significant decreases through at least 2 weeks after the surgery (P > 0.05). There was greater cerebral blood volume and flow in the GBM than in the normal-appearing white matter (1.46 ± 0.25 vs. 1.13 ± 0.16 and 1.30 ± 0.22 vs. 1.02 ± 0.14; P < 0.001 and P < 0.001, respectively). The MVD in the histologic specimens was correlated with the cerebral blood volume in the GBM tissue (r = 0.850, P = 0.004).

CONCLUSION

Our results suggest that the canine GBM model showed perfusion imaging characteristics similar to those of humans, and it might have potential as a model to assess novel technical developments for GBM treatment.

Radioresistance in Glioblastoma and the Development of Radiosensitizers

Ionizing radiation is a common and effective therapeutic option for the treatment of glioblastoma (GBM). Unfortunately, some GBMs are relatively radioresistant and patients have worse outcomes after radiation treatment. The mechanisms underlying intrinsic radioresistance in GBM has been rigorously investigated over the past several years, but the complex interaction of the cellular molecules and signaling pathways involved in radioresistance remains incompletely defined. A clinically effective radiosensitizer that overcomes radioresistance has yet to be identified. In this review, we discuss the current status of radiation treatment in GBM, including advances in imaging techniques that have facilitated more accurate diagnosis, and the identified mechanisms of GBM radioresistance. In addition, we provide a summary of the candidate GBM radiosensitizers being investigated, including an update of subjects enrolled in clinical trials. Overall, this review highlights the importance of understanding the mechanisms of GBM radioresistance to facilitate the development of effective radiosensitizers.

Computed Tomography and Magnetic Resonance Imaging Are Equivalent in Mensuration and Similarly Inaccurate in Grade and Type Predictability of Canine Intracranial Gliomas

While magnetic resonance imaging (MRI) is the gold-standard imaging modality for diagnosis of intracranial neoplasia, computed tomography (CT) remains commonly used for diagnosis and therapeutic planning in veterinary medicine. Despite the routine use of both imaging modalities, comparison of CT and MRI has not been described in the canine patient. A retrospective study was performed to evaluate CT and MRI studies of 15 dogs with histologically confirmed glioma. Multiple lesion measurements were obtained, including two-dimensional and volumetric dimensions in pre-contrast and post-contrast images. Similar measurement techniques were compared between CT and MRI. The glioma type (astrocytoma or oligodendroglioma) and grade (high or low) were predicted on CT and MRI independently. With the exception of the comparison between CT pre-contrast volume to T2-weighted MRI volume, no other statistical differences between CT and MRI measurements were identified. Overall accuracy for tumor grade (high or low) was 46.7 and 53.3% for CT and MRI, respectively. For predicted tumor type, accuracy of CT was 53.3% and MRI and MRI 60%. Based on the results of this study, both CT and MRI contrast measurement techniques are considered equivalent options for lesion mensuration. Given the low-to-moderate predictability of CT and MRI in glioma diagnosis, histopathology remains necessary for accurate diagnosis of canine brain tumors.

Dogs are man's best friend: in sickness and in health

With the median survival of 14.6 months following best available standard of care, malignant gliomas (MGs) remain one of the biggest therapeutic challenges of the modern time. Although the last several decades have witnessed tremendous advancement in our understanding of MG and evolution of many successful preclinical therapeutic strategies, even the most successful preclinical therapeutic strategies often fail to cross the phase I/II clinical trial threshold. One of the significant, but less commonly discussed, barriers in developing effective glioma therapy is the lack of a robust preclinical model. For the last 30 years, rodent orthotopic xenograft models have been extensively used in the preclinical setting. Although they provide a good basic model for understanding tumor biology, their value in successfully translating preclinical therapeutic triumph into clinical success is extremely poor. Companion dogs, which share the same environmental stress as their human counterparts, also spontaneously develop MGs. Dog gliomas that develop spontaneously in an immunocompetent host are very similar to human gliomas and potentially provide a stronger platform for validating the efficacy of therapeutic strategies proven successful in preclinical mouse models. Integrating this model can accelerate development of effective therapeutic options that will benefit both human subjects and pet dogs.

New treatment modalities for brain tumors in dogs and cats

Despite advancements in standard therapies, intracranial tumors remain a significant source of morbidity and mortality in veterinary and human medicine. Several newer approaches are gaining more widespread acceptance or are currently being prepared for translation from experimental to routine therapeutic use. Clinical trials in dogs with spontaneous brain tumors have contributed to the development and human translation of several novel therapeutic brain tumor approaches.

Invited review--neuroimaging response assessment criteria for brain tumors in veterinary patients

The evaluation of therapeutic response using cross-sectional imaging techniques, particularly gadolinium-enhanced MRI, is an integral part of the clinical management of brain tumors in veterinary patients. Spontaneous canine brain tumors are increasingly recognized and utilized as a translational model for the study of human brain tumors. However, no standardized neuroimaging response assessment criteria have been formulated for use in veterinary clinical trials. Previous studies have found that the pathophysiologic features inherent to brain tumors and the surrounding brain complicate the use of the response evaluation criteria in solid tumors (RECIST) assessment system. Objectives of this review are to describe strengths and limitations of published imaging-based brain tumor response criteria and propose a system for use in veterinary patients. The widely used human Macdonald and response assessment in neuro-oncology (RANO) criteria are reviewed and described as to how they can be applied to veterinary brain tumors. Discussion points will include current challenges associated with the interpretation of brain tumor therapeutic responses such as imaging pseudophenomena and treatment-induced necrosis, and how advancements in perfusion imaging, positron emission tomography, and magnetic resonance spectroscopy have shown promise in differentiating tumor progression from therapy-induced changes. Finally, although objective endpoints such as MR imaging and survival estimates will likely continue to comprise the foundations for outcome measures in veterinary brain tumor clinical trials, we propose that in order to provide a more relevant therapeutic response metric for veterinary patients, composite response systems should be formulated and validated that combine imaging and clinical assessment criteria.

Vertebrate animal models of glioma: understanding the mechanisms and developing new therapies

Glioblastoma Multiforme (GBM) is recognized as one of the most deadly cancers characterized by cellular atypia, severe necrosis, and high rate of angiogenesis. In this review, we discuss a diversified group of GBM xenograft models and compare them with the genetically engineered mouse (GEM) model systems. Next, we describe common genetic defects observed in GBM and numerous GEM models that recapitulate these abnormalities. Finally, we focus on the clinical value of other vertebrate animal models such as the canine model by examining their contributions to GBM research.

Supratentorial low-grade glioma resectability: statistical predictive analysis based on anatomic MR features and tumor characteristics

PURPOSE

To retrospectively assess the main variables that affect the complete magnetic resonance (MR) imaging-guided resection of supratentorial low-grade gliomas.

MATERIALS AND METHODS

Institutional review board approval was obtained for this retrospective HIPAA-compliant study, with the requirement for informed consent waived. Data from 101 patients (61 men, 40 women; mean age, 39 years; age range, 18-72 years) who had nonenhancing supratentorial mass lesions that were histopathologically diagnosed as low-grade (World Health Organization grade II) gliomas and consecutively underwent surgery with intraoperative MR imaging guidance were analyzed. There were 21 low-grade astrocytomas, 64 oligodendrogliomas, and 16 mixed oligoastrocytomas. Initial and residual tumor volumes were measured on intraoperative T2-weighted MR images and three-dimensional spoiled gradient-echo MR images. The anatomic relationships between the tumor and eloquent cortical and/or subcortical regions and the influence of these relationships on the extent of resection were analyzed on the basis of preoperative MR imaging findings. Summary measures, univariate Fisher exact test and t test, and multivariate logistic regression analyses were performed.

RESULTS

Tumor volume ranged from 2.7-231.0 mL. Univariate analyses revealed the following tumor characteristics to be significant predictive variables of incomplete tumor resection: diffuse tumor margin on T2-weighted MR images, oligodendroglioma or oligoastrocytoma histopathologic type, and large tumor volume (P < .05 for all). Tumor involvement of the following structures was associated with incomplete resection: corpus callosum, corticospinal tract, insular lobe, middle cerebral artery, motor cortex, optic radiation, visual cortex, and basal ganglia (P < .05 for all). Multivariate analyses revealed that incomplete tumor resection was due to tumor involvement of the corticospinal tract (P < .01), large tumor volume (P < .01), and oligodendroglioma histopathologic type (P = .02).

CONCLUSION

The main variables associated with incomplete tumor resection in 101 patients were identified by using statistical predictive analyses.

Cavernous sinus tumor model in the canine: a simulation model for cavernous sinus tumor surgery

OBJECTIVE

The recent limitations of working hours for neurosurgical trainees carry the risk of decreasing the amount of microsurgical experience. In the absence of enough surgical exposure to some pathological states, an alternative option of a more continuous source of tactile and visual experience that simulates the real-life state is needed. To help with this problem, we established a cavernous sinus tumor model in the canine.

METHODS

A gliosarcoma cell line that was harvested from a tumor model in nude mice was implanted in six mongrel dogs. In the first group (two dogs), the cell line was implanted in the dural leaflets of the cavernous sinus. (Immunosuppression was used in one dog.) In the second group (four dogs), the cell line was implanted in the region of the gasserian ganglion. (Immunosuppression was used in all four dogs.) The condition of each dog was followed through neurological examinations and serial magnetic resonance imaging. The cavernous sinus region later was explored, after which the dogs were later killed and histopathological evaluations of the cavernous sinus region was carried out.

RESULTS

The initial cell line implanted within the dural leaflets of the cavernous sinus showed no evidence of tumor growth. The tumor grew in all four dogs that had the gliosarcoma cell line implanted in the region of the gasserian ganglion. The clinical and radiological features as well as the experience of the surgical dissection of these tumors simulated cavernous sinus tumors in humans.

CONCLUSION

We established the first cavernous sinus tumor model in the canine. This model simulates the real-life pathological state, and it can be used as an alternative source of surgical experience to advance surgical skills.

The importance of routine magnetic resonance imaging in trigeminal neuralgia diagnosis

The aim of the study was to determine the frequency of structural lesions diagnosed on magnetic resonance imaging (MRI) in a sample of patients with trigeminal neuralgia (TN) and to compare history and clinical features between the groups with and without structural lesions. Clinical records and MRI findings of 42 consecutive patients diagnosed with TN at the National Dental Centre, Singapore, and who underwent routine MRI examinations at diagnostic workup between April 1997 and March 1999 were retrospectively studied. Of these, 6 (14.3%; 95% confidence interval, 0 to 28.8%) were diagnosed with an associated structural lesion. Mean age of the group with structural lesions was 53.3 years (standard deviation, 10.9 years) at presentation, there was female predominance (M:F = 1:2), and all (100.0%) gave a typical pain history and had unilateral involvement of a single division of the trigeminal nerve. Two patients (33.3%) had other cranial nerve abnormalities. Three patients (50.0%) had complete, 2 patients (33.3%) had partial, and 1 patient (16.7%) had poor response to medical therapy. None of these variables was statistically different from the group without structural lesions (cranial nerve abnormalities were not compared). In conclusion, the frequency of structural lesions associated with TN in our sample was relatively high. It is not possible to reliably identify high risk patients for selective MRI indication on the basis of history and clinical features. It may be prudent to consider routine MRI for all patients with TN to exclude structural lesions.

Determination of intracranial tumor volumes in a rodent brain using magnetic resonance imaging, Evans blue, and histology: a comparative study

The measurement of tumor volumes is a practical and objective method of assessing the efficacy of a therapeutic agent. However, the relative accuracy of different methods of assessing tumor volume has been unclear. Using T1-weighted, gadolinium-enhanced magnetic resonance Imaging (T1-MRI), Evans Blue infusion and histology we measured intracranial tumor volumes in a rodent brain tumor model (RT2) at days 10, 16 and 18 after implantation of cells in the caudate putamen. There is a good correlation between tumor volumes comparing T1-MRI and Evans Blue (r2 = 0.99), T1-MRI and Histology (r2 = 0.98) and histology and Evans Blue (r2 = 0.93). Each of these methods is reliable in estimating tumor volumes in laboratory animals. There was significant uptake of gadolinium and Evans Blue in the tumor suggesting a wide disruption of the blood-brain barrier.

Preclinical evaluation of benzoporphyrin derivative combined with a light-emitting diode array for photodynamic therapy of brain tumors

OBJECTIVE

The aim of this study was to investigate the second-generation photosensitizer benzoporphyrin derivative (BPD) and a novel light source applicator based on light-emitting diode (LED) technology for photodynamic therapy (PDT) of brain tumors.

METHODS

We used a canine model to investigate normal brain stem toxicity. Twenty-one canines underwent posterior fossa craniectomies followed by PDT with BPD. These animals were compared to light only and BPD control. In addition, we investigated the ability of BPD and LED to cause inhibition of cell growth in canine glioma and human glioma cell lines, in vitro. The biodistribution of BPD labeled with 111In-BPD in mice with subcutaneous and intracerebral gliomas and canines with brain tumors was studied.

RESULTS

The in vivo canine study resulted in a maximal tolerated dose of 0.75 mg/kg of BPD and 100 J/cm(2) of LED light for normal brain tissue. The in vitro study demonstrated 50% growth inhibition for canine and human glioma cell lines of 10 and 4 ng/ml, respectively. The mucine study using 111In-BPD showed a tumor to normal tissue ratio of 12:1 for intracerebral tumors and 3.3:1 for subcutaneous tumors. Nuclear scans of canines with brain tumors showed uptake into tumors to be maximal from 3 to 5 h.

CONCLUSION

Our study supports that BPD and LED light sources when used at appropriate drug and light doses limit normal brain tissue toxicity at doses that can cause significant glioma cell toxicity in vitro. In addition, there is higher BPD uptake in brain tumors as compared to normal brain in a mouse glioma model. These findings make BPD a potential new-generation photosensitizer for the treatment of childhood posterior fossa tumors as well as other malignant cerebral pathology.

Cerebral gliomas and metastases: assessment with contrast-enhanced fast fluid-attenuated inversion-recovery MR imaging

The effect of contrast material on fast fluid-attenuated inversion-recovery (FLAIR) magnetic resonance images was evaluated for 16 patients with enhancing gliomas and 12 patients with cerebral metastases. Because of a marked T1 effect, fast FLAIR imaging provided a marked contrast enhancement, resulting in the highest tumor-to-background contrast ratio compared with standard imaging techniques.

Intraoperative magnetic resonance imaging with the magnetom open scanner: concepts, neurosurgical indications, and procedures: a preliminary report

OBJECTIVE

Intraoperative magnetic resonance imaging (MRI) is now available with the General Electric MRI system for dedicated intraoperative use. Alternatively, non-dedicated MRI systems require fewer specific adaptations of instrumentation and surgical techniques. In this report, clinical experiences with such a system are presented.

METHODS

All patients were surgically treated in a "twin operating theater," consisting of a conventional operating theater with complete neuronavigation equipment (StealthStation and MKM), which allowed surgery with magnetically incompatible instruments, conventional instrumentation and operating microscope, and a radiofrequency-shielded operating room designed for use with an intraoperative MRI scanner (Magnetom Open; Siemens AG, Erlangen, Germany). The Magnetom Open is a 0.2-T MRI scanner with a resistive magnet and specific adaptations that are necessary to integrate the scanner into the surgical environment. The operating theaters lie close together, and patients can be intraoperatively transported from one room to the other. This retrospective analysis includes 55 patients with cerebral lesions, all of whom were surgically treated between March 1996 and September 1997.

RESULTS

Thirty-one patients with supratentorial tumors were surgically treated (with navigational guidance) in the conventional operating room, with intraoperative MRI for resection control. For 5 of these 31 patients, intraoperative resection control revealed significant tumor remnants, which led to further tumor resection guided by the information provided by intraoperative MRI. Intraoperative MRI resection control was performed in 18 transsphenoidal operations. In cases with suspected tumor remnants, the surgeon reexplored the sellar region; additional tumor tissue was removed in three of five cases. Follow-up scans were obtained for all patients 1 week and 2 to 3 months after surgery. For 14 of the 18 patients, the images obtained intraoperatively were comparable to those obtained after 2 to 3 months. Intraoperative MRI was also used for six patients undergoing temporal lobe resections for treatment of pharmacoresistant seizures. For these patients, the extent of neocortical and mesial resection was tailored to fit the preoperative findings of morphological and electrophysiological alterations, as well as intraoperative electrocorticographic findings.

CONCLUSION

Intraoperative MRI with the Magnetom Open provides considerable additional information to optimize resection during surgical treatment of supratentorial tumors, pituitary adenomas, and epilepsy. The twin operating theater is a true alternative to a dedicated MRI system. Additional efforts are necessary to improve patient transportation time and instrument guidance within the scanner.

A comparison between preoperative magnetic resonance and intraoperative ultrasound tumor volumes and margins

A major obstacle in surgical neuro-oncology is differentiating the interface between tumor and normal brain. Twenty-two brain tumors were evaluated preoperatively with magnetic resonance imaging. Intraoperative ultrasonography was used to guide surgical resection of these tumors, and results were compared with surgical and pathologic findings. Ultrasound tumor volume estimates were larger than T1 gadolinium-enhanced and T1 non-gadolinium-enhanced volumes, but these differences did not reach statistical significance. Similarly T2 volumes were larger than the corresponding sonographic volumes, except for the subset of low-grade gliomas, and in that instance the difference was small, but again the differences were not statistically significant. Ultrasonography enhanced identification of infiltrating tumor cells beyond falsely underestimated tumor margins as defined by T1 images. Ultrasound images helped differentiate edema as seen on T2 images from solid tumor and normal brain. The information gained from ultrasound images can be used to enhance tumor resection and improve patient survival and quality of life.

Early postoperative magnetic resonance imaging after resection of malignant glioma: objective evaluation of residual tumor and its influence on regrowth and prognosis

In the vast majority of studies that address the role of surgery in the management of high-grade gliomas, the degree of tumor removal accomplished is solely based on the intraoperative perception of the neurosurgeon. Despite its fundamental importance for a comparison of different treatment modalities, little systematic effort has been made to evaluate the residual gross tumor by neuroimaging methods immediately after surgery. We report the results of a prospective study using contrast-enhanced computed tomography and magnetic resonance imaging (MRI) to monitor 60 patients after the resection of a high-grade glioma. In each case, the first scans were obtained between Days 1 and 5 after surgery, followed by serial imaging every 2 to 3 months, usually until the condition of the patient deteriorated severely or the patient died. Gadolinium-enhanced MRI proved to be extremely valuable for assessing gross residual tumor when performed during Days 1 to 3 after the resection of a preoperatively enhancing high-grade glioma. This timing avoided surgically induced contrast enhancement and minimized interpretative difficulties. In delineating residual tumor, MRI was vastly superior to computed tomography. About 80% of tumor "recurrences" emerged from definitely enhancing remnants, as revealed by early postoperative MRI. The neurosurgeon's estimation of gross tumor burden reduction could be shown to be much less accurate (by a factor of 3) than the postoperative assessment by modern neuroimaging. In our series, residual tumor enhancement was the most predictive prognostic factor of survival in patients with glioblastoma, followed by radiotherapy. Patients with a residual tumor postoperatively had a 6.595-times higher risk of death in comparison to patients without a residual tumor. Patients undergoing radiotherapy had a 0.258-times lower risk of death in comparison to patients who were not treated with radiation. Concerning survival, the prognostic significance of both variables surpassed age and performance.