Dynamic contrast-enhanced perfusion processing for neuroradiologists: model-dependent analysis may not be necessary for determining recurrent high-grade glioma versus treatment effect.

BACKGROUND AND PURPOSE: Dynamic contrast-enhanced perfusion MR imaging has proved useful in determining whether a contrast-enhancing lesion is secondary to recurrent glial tumor or is treatment-related. In this article, we explore the best method for dynamic contrast-enhanced data analysis. MATERIALS AND METHODS: We retrospectively reviewed 24 patients who met the following conditions: 1) had at least an initial treatment of a glioma, 2) underwent a half-dose contrast agent (0.05-mmol/kg) diagnostic-quality dynamic contrast-enhanced perfusion study for an enhancing lesion, and 3) had a diagnosis by pathology within 30 days of imaging. The dynamic contrast-enhanced data were processed by using model-dependent analysis (nordicICE) using a 2-compartment model and model-independent signal intensity with time. Multiple methods of determining the vascular input function and numerous perfusion parameters were tested in comparison with a pathologic diagnosis. RESULTS: The best accuracy (88%) with good correlation compared with pathology (P = .005) was obtained by using a novel, model-independent signal-intensity measurement derived from a brief integration beginning after the initial washout and by using the vascular input function from the superior sagittal sinus for normalization. Modeled parameters, such as mean endothelial transfer constant > 0.05 minutes(-1), correlated (P = .002) but did not reach a diagnostic accuracy equivalent to the model-independent parameter. CONCLUSIONS: A novel model-independent dynamic contrast-enhanced analysis method showed diagnostic equivalency to more complex model-dependent methods. Having a brief integration after the first pass of contrast may diminish the effects of partial volume macroscopic vessels and slow progressive enhancement characteristic of necrosis. The simple modeling is technique- and observer-dependent but is less time-consuming.

Post-epidural analgesia spinal cord infarction: MRI correlation.

Permanent neurological deficits after epidural analgesia are rare, but have long been believed to be caused by cord ischaemia when no obvious cause is demonstrable. The mechanisms of this injury are uncertain, but a literature review suggests important risk factors. We report a first case of extensive spinal cord infarction confirmed by magnetic resonance imaging (MRI) following post-thoracotomy epidural analgesia and review the literature to explain the mechanism underlying this devastating complication.

Choroid plexus cyst and chordoid glioma. Report of two cases.

Several types of mass lesions may occur in the third and lateral ventricles. Typically they arise from the lining of the ventricular cavity or from contiguous structures, by extension into the ventricle. The authors describe two patients, each of whom presented with a different rare lesion of the ventricular system. The first was a 53-year-old woman with a history of hypertension who sustained a blunt traumatic injury to the occipital region and subsequently developed a progressively worsening right-sided headache. Radiological examinations over the next 2 years revealed an enlarged right lateral ventricle and, ultimately, a choroid plexus cyst in its anterior and middle third, near the foramen of Monro, which is a rare location for these lesions. The cyst was removed en bloc, and follow-up examinations showed a significant improvement in her headache and a minimal differences in size between right and left ventricles. The authors also describe a 57-year-old man with hypertension, diabetes mellitus, and an old mycardial infarct, who presented to an outside institution with a progressively worsening headache, generalized malaise, and loss of olfactory sensation. Diagnostic imaging revealed a 1.5-cm oval lesion centered in the lamina terminalis region, an open craniotomy was performed, and evaluation of a biopsy sample demonstrated the mass to be a chordoid glioma of the third ventricle, a recently described glioma subtype. Two days after surgery, he suffered a left parietal stroke and an anterior mycardial infarction. After convalescing, he presented to The University of Texas M. D. Anderson Cancer Center for radiotherapy and follow up; 7 months later he was readmitted complaining of headache, short-term memory loss, and worsening confusion and disorientation. Neuroimaging revealed progression of the tumor (now 2 cm in diameter), which was removed by gross-total resection. His headache resolved immediately, and 2 months later his only complaint was of episodes of confusion. Three weeks later he died of a massive myocardial infarction. These two patients represent the sixth case of an adult with a choroid plexus cyst in the anterior lateral ventricle and the 19th case of an adult with a chordoid glioma of the third ventricle, respectively.

Evolution of diagnostic neuroradiology from 1904 to 1999.

Neuroradiology began in the early 1900s soon after Roentgen discovered x rays, with the use of skull radiographs to evaluate brain tumors. This was followed by the development of ventriculography in 1918, pneumoencephalography in 1919, and arteriography in 1927. In the beginning, air studies were the primary modality, but this technique was supplanted by angiography in the 1950s and 1960s. The first full-time neuroradiologist in the United States was Cornelius G. Dyke at the New York Neurological Institute in 1930. Neuroradiology took a firm hold as a specialty in the early 1960s when Dr Juan M. Taveras brought together fourteen neuroradiologists from the United States and Canada to establish the nucleus of what was to become the American Society of Neuroradiology, or ASNR. This society's initial goals were to perform research and to advance knowledge within the specialty. Neuroradiologists initially were able to diagnose vascular disease, infections, tumors, trauma, and alterations in cerebrospinal fluid flow, because the brain structure was invisible. Neuroradiology was forever changed with computed tomography (CT) because the brain structure became visible. Soon thereafter, magnetic resonance (MR) imaging was developed, and it not only provided anatomic but also made possible vascular and physiologic functional imaging.

Malignant gliomas: MR imaging spectrum of radiation therapy- and chemotherapy-induced necrosis of the brain after treatment.

PURPOSE: To describe both the common and less frequently encountered magnetic resonance (MR) imaging features of radiation therapy- and chemotherapy-induced brain injury, with particular emphasis on radiation necrosis. MATERIALS AND METHODS: A cohort of 148 adult patients underwent surgical resection of malignant brain (glial) tumors and were subsequently entered into a research protocol that consisted of accelerated radiation therapy with carboplatin followed by chemotherapy with procarbazine, lomustine, and vincristine. Patients typically underwent sequential MR imaging at 6-8-week intervals during the 1st year and at 3-6-month intervals during subsequent years. In all patients, histopathologic confirmation of lesion composition was performed by board-certified neuropathologists. RESULTS: The patients exhibited different types of MR imaging-detected abnormalities of the brain: pure radiation necrosis in 20 patients, a mixture of predominantly radiation necrosis with limited recurrent and/or residual tumor (less than 20% of resected tissue) in 16 patients, radiation necrosis of the cranial nerves and/or their pathways in two patients, radiation-induced enhancement of the white matter in 52 patients, and radiation-induced enhancement of the cortex in nine patients. CONCLUSION: The frequent diagnostic dilemma of recurrent neoplasm versus radiation necrosis is addressed in this study through a description of the varying spatial and temporal patterns of radiation necrosis at MR imaging.

Intracranial leptomeningeal metastases: comparison of depiction at FLAIR and contrast-enhanced MR imaging.

PURPOSE: To compare contrast material-enhanced T1-weighted and fluid-attenuated inversion-recovery (FLAIR) magnetic resonance (MR) images in depicting leptomeningeal metastases. MATERIALS AND METHODS: Malignant lesions detected at cytologic examination of cerebrospinal fluid in 70 patients were reviewed. There were 58 studies in which both FLAIR and contrast-enhanced T1-weighted spin-echo MR images were available. A senior neuroradiologist reviewed the images from each sequence individually and separately for signs of leptomeningeal metastases and assigned a diagnostic rating of positive, indeterminate, or negative. RESULTS: Leptomeningeal metastases were depicted in 38 cases on contrast-enhanced T1-weighted spin-echo images and in 20 cases on FLAIR images. In three cases, leptomeningeal metastases were detected by using only FLAIR images. In 20 cases, leptomeningeal metastases were detected by using only contrast-enhanced T1-weighted spin-echo images. FLAIR imaging has a sensitivity of 34% for cytologically proved leptomeningeal metastases. Gadolinium-enhanced MR imaging has a sensitivity of 66%. CONCLUSION: Used alone, contrast-enhanced T1-weighted images are better than FLAIR images for detecting leptomeningeal metastases. This is particularly true for cases in which leptomeningeal metastases manifest primarily or solely as cranial nerve involvement.

Quantitative morphology of the corpus callosum in children with neurofibromatosis and attention-deficit hyperactivity disorder.

Neurofibromatosis-1 is a common autosomal-dominant genetic disorder associated with numerous physical anomalies and an increased incidence of attention-deficit hyperactivity disorder (ADHD). Studies of children with idiopathic ADHD have suggested a link between corpus callosum size and symptom severity. This study examines the contribution of corpus callosum morphology to symptoms of ADHD in children with neurofibromatosis. Eighteen control subjects and 36 children with neurofibromatosis underwent magnetic resonance imaging of the brain. Twelve subjects with neurofibromatosis had evidence of ADHD and 24 did not. Subjects with neurofibromatosis had significantly larger total corpus callosum area and significantly larger regional measurements in three of seven areas. However, there were no differences between the neurofibromatosis alone and neurofibromatosis plus ADHD groups. Increased severity of attention problems was associated with smaller total callosal areas. These results suggest that some features of ADHD in children with neurofibromatosis could be linked to quantifiable differences in brain morphology, but the nature of the genetic mutation in neurofibromatosis suggests that neurochemical effects also could be important.

Brain volume in children with neurofibromatosis type 1: relation to neuropsychological status.

OBJECTIVE: To determine characteristics of brain morphology in children and adolescents with neurofibromatosis type 1 and relate these characteristics to neuropsychological functioning. BACKGROUND: Neurofibromatosis type 1 is associated with numerous CNS abnormalities and cognitive impairment. Abnormal high signal intensity visible on brain MRI, brain tumors, and macrocephaly are common. Research into links between neuroanatomic and cognitive features has been inconclusive. METHODS: Fifty-two children and adolescents with neurofibromatosis type 1 were compared with 19 control subjects on several quantitative neuroanatomic and neuropsychological measures. RESULTS: Total brain volume, especially gray matter, was significantly greater for neurofibromatosis type 1 subjects than the control subjects. Group differences in the ratio of gray matter to white matter were more prominent in younger than in older subjects. Volume of gray matter in the subjects with neurofibromatosis type 1 was related to their degree of learning disability. Corpus callosum size was significantly larger for subjects in the neurofibromatosis type 1 group, and diminished performance on measures of academic achievement and visual-spatial and motor skills were associated with greater regional corpus callosum size. CONCLUSIONS: Neuroanatomic morphology and the developmental pattern of gray matter and white matter in subjects with neurofibromatosis type 1 differed from in control subjects. Some of these differences are related to the neuropsychological status of the neurofibromatosis type 1 group. We propose that delayed developmental apoptosis results in macrocephaly and a delay in the development of appropriate neuronal connections in children with neurofibromatosis type 1. We further propose that these morphologic delays are related to the cognitive profile of neurofibromatosis type 1.

Magnetic resonance imaging of benign spinal lesions simulating metastasis: role of diffusion-weighted imaging.

In an attempt to avoid unnecessary therapy, this article demonstrates benign vertebral body lesions that mimic metastatic disease in cancer patients with back pain. The magnetic resonance imaging features that aid in differential diagnosis are demonstrated. In addition, the value of diffusion-weighted spinal imaging to further aid in distinguishing benign from malignant disease is described.

Correlation between dynamic MRI and outcome in patients with malignant gliomas.

We assessed the correlation between dynamic MRI results and clinical outcomes in patients with malignant gliomas. Rapid serial MRIs were obtained after bolus injection of gadolinium that resulted in an initial fast uptake followed by a slow uptake of contrast. The maximum rate of uptake and delayed rate of uptake were correlated with survival and prognostic covariates such as age and histology. In 121 subjects, higher maximum uptake rates, 3.6 signal intensity units per second or greater, were associated with shorter survival (p = 0.0066). The correlation of delayed rate of uptake with survival was less significant. After adjusting for age, histology, and Karnofsky performance score, the maximum rate of uptake remained more significantly correlated with survival than the delayed rate of uptake. Thirty-one patients had surgery within 1 month of dynamic MRI, and those with glioblastoma multiforme or anaplastic gliomas had higher maximum rates of uptake than those with pure necrosis or mixed tumor and necrosis (p = 0.022). No correlation between delayed rate of uptake and histology was seen in this group of patients. Our results suggest that the maximum rate of uptake in dynamic MRI can be a prognostic measure for patients with malignant gliomas. Further prospective study is needed to assess the utility of this technique for evaluating brain tumors.

The significance of lack of MR contrast enhancement of supratentorial brain tumors in adults: histopathological evaluation of a series.

BACKGROUND: To correlate magnetic resonance imaging (MRI) findings of non-enhancement of supratentorial brain neoplasms in adults with histopathologic findings. METHODS: Forty adult patients whose preoperative MRI studies demonstrated a non-enhancing supratentorial brain neoplasm were identified retrospectively. Biopsy material for all patients was then reviewed by a board-certified neuropathologist. RESULTS: Histopathologic examination identified 24 (60%) low-grade gliomas: 4 (10%) low-grade astrocytomas, 10 (25%) low-grade gliomas (not further classified), 8 (20%) low-grade oligodendrogliomas, and 2 (5%) low-grade mixed oligoastrocytomas. However, 16 (40%) nonenhancing lesions were classified as anaplastic gliomas: 12 (30%) anaplastic astrocytomas, 1 (2.5%) anaplastic mixed oligoastrocytoma, 1 (2.5%) anaplastic oligodendroglioma, and 2 (5%) anaplastic infiltrating gliomas of indeterminate subtype. CONCLUSION: Non-enhancement of supratentorial brain neoplasms in adults does not equate with low-grade malignancy. This fact should be taken into account when biopsy and treatment are being planned in patients with nonenhancing brain tumors. More aggressive and/or surgical therapy might be indicated for such lesions, particularly those in the nondominant hemisphere or nonmotor areas.

Dynamic imaging of intracranial lesions using fast spin-echo imaging: differentiation of brain tumors and treatment effects.

The purpose of this study was to develop a technique for differentiating between recurrent brain tumors and treatment-related changes, such as radiation necrosis, using dynamic MRI. Ninety-five patients with intracranial mass lesions were evaluated using T1-weighted fast spin-echo (FSE) MRI at 1.5 T. Pathologies included treatment-related changes (n = 32), primary tumors (n = 41), metastatic tumors (n = 5), meningiomas (n = 4), and mixed primary/treatment related changes (n = 13). Signal enhancement-time curves were analyzed by fitting to a sigmoidal-exponential function. Maximal enhancement rates were calculated as the first derivative of the fitted curve. Based on the maximal enhancement rates, treatment-related changes could be differentiated from primary tumors, metastatic tumors, and meningiomas at the P < .05 confidence level. Lesions of mixed tumor and treatment-related change had intermediate values. Dynamic MRI can be used to differentiate treatment-related changes from primary tumors in previously treated patient populations based on maximal enhancement rates. Individual case studies demonstrate the clinical significance of these findings.

A second look at unenhanced spinal magnetic resonance imaging of malignant leptomeningeal disease.

The purpose of this study is to evaluate the use of unenhanced magnetic resonance (MR) imaging in identifying malignant leptomeningeal disease (MLD). Included in this study were fifty patients with evidence of leptomeningeal enhancement on post-gadolinium MR images and with cytological confirmation of MLD. Unenhanced, T1-weighted spin-echo (SE) MR images of the spine were analyzed for loss of cerebrospinal fluid (CSF) clarity, poor definition of the conus medullaris, thickened and clumped nerve roots, and nodules. Patterns of leptomeningeal enhancement on post-gadolinium, T1-weighted SE images were noted. Findings of MLD on unenhanced MR images were observed in 41 (85%) of 48 studies of the lumbar spine, 10 (50%) of 20 studies of the thoracic spine, and two (33%) of six studies of the cervical spine. In the lumbar spine, thickened and clumped nerve roots, poor definition of the conus medullaris, loss of CSF clarity, and nodules were observed with decreasing frequency. The signs of MLD on unenhanced images of the cervicothoracic spine included nodules and clouding of CSF. Patterns of leptomeningeal enhancement included linear, linear/nodular, nodular, enhancement of nerve roots, and stacking, with tumor filling the lumbosacral canal. Findings of MLD were present on 73% of the unenhanced MR images of the spine. Recognition of MLD on unenhanced MR images can guide the appropriate work-up and therapeutic approach.

Magnetic resonance imaging of the central nervous system of the rhesus monkey.

The purpose of the study reported here was to document the ability of magnetic resonance imaging to depict the imaging characteristics of normal structures within the central nervous system of adult rhesus monkeys. The head and the cervical and thoracic parts of the spinal cord of two rhesus monkeys were imaged in a clinical 1.5-T whole-body imager. Specific images were selected, and some notable structures were identified. Results of this study document the usefulness of MRI as an expeditious, noninvasive research and diagnostic imaging technique and illustrates the normal magnetic resonance signal patterns of the brain and spinal cord in rhesus monkeys.

A review of MRI pulse sequences and techniques in neuroimaging.

BACKGROUND: The unmatched soft tissue contrast provided by magnetic resonance imaging (MRI) has made it the modality of choice for many neuroimaging examinations. The fact that signal intensity in MRI depends on many parameters, including spin-lattice and spin-spin relaxation times, proton density, and velocity, makes it possible to highlight various pathologies by appropriate choice of pulse sequences and pulse sequence parameters. It is somewhat overwhelming however, to filter through various pulse sequences and parameters in order to understand how their selection affects image contrast. This brief review is intended to highlight common pulse sequences and parameters as well as introduce new techniques currently being released for clinical use. MATERIALS: Basic pulse sequences are described and the influence of the acquisition parameters on image contrast are illustrated. Such basic sequences include the ubiquitous spin echo, fast spin echo, and gradient echo sequences. Specialized techniques for fat suppression and magnetic resonance angiography are also presented. Currently approved contrast agents for use in MRI are briefly reviewed, and various advanced pulse sequences, such as those for diffusion and magnetization transfer contrast imaging, are briefly outlined. RESULTS: The utility of basic and advanced pulse sequences are demonstrated by clinical examples and images of normal brain and spine. New sequences and techniques are briefly outlined with regard to their potential for improving neuroimaging examinations. CONCLUSIONS: This brief review outlines how the choice of pulse sequence and acquisition parameters influences the resulting image contrast for a variety of basic and advanced imaging techniques.

MR prediction of benign and malignant vertebral compression fractures.

We reviewed spinal MR images of 58 patients with 98 compressed vertebrae. Benign (47 vertebrae) or malignant (51 vertebrae) etiology was established by biopsy or radiologic follow-up. Compressed vertebrae were analyzed for presence and characteristics of signal abnormality, altered vertebral contour, Schmorl's nodes, pedicular involvement, and contrast uptake. Statistical analysis was performed. Diffuse and homogeneous decrease in signal intensity on T1-weighted images, convex vertebral contour, involvement of the pedicles, and a lumbar location were more frequently observed in malignant fractures (P < .01). A thoracic location, lack of signal change, or a band-like abnormality and absence of pedicular involvement or contour abnormality characterized benign fractures (P < .01). Schmorl's nodes and enhancement did not help establish a diagnosis. When a constellation of MR criteria are applied, the accuracy of the diagnosis of malignant and benign vertebral compression fractures may reach 94%.

Glioblastoma multiforme arising in the irradiated spinal cord of a rhesus monkey (Macaca mulatta).

An adult female rhesus monkey that had received 44.0 Gy of cobalt 60 radiation to 8 cm of the cervical and upper thoracic spinal cord approximately 2.8 years postirradiation developed a sudden onset of self-mutilation and loss of function of the right arm followed progressively by loss of function of the left arm and terminally bilateral paresis of the legs. Histopathologic examination of the cervical spinal cord revealed a glioblastoma multiforme that extended from the cervical medullary junction to the sixth cervical vertebrae. Because of the infrequent occurrence of spontaneous neoplasia in rhesus monkeys and the location in the radiation field, the glioblastoma is believed to be radiation induced.

Multiple brain metastases: a rare manifestation of adenoid cystic carcinoma of the parotid gland.

A rare case of multiple hematogenous brain metastases from adenoid cystic carcinoma of the parotid gland is reported. The patient had a parotid tumor that was treated ten years prior to the appearance of the brain metastases. Magnetic resonance imaging (MRI) and histological findings, as well as the radiation therapy response, of this tumor are presented.

Prognostic significance of preoperative MRI scans in glioblastoma multiforme.

Tumor necrosis, enhancement, and associated edema in patients with glioblastoma multiforme (GBM) represent biological variables that can be quantitated on preoperative MRI scans. We reviewed 48 highly selected patients, all of whom had supratentorial lesions, had undergone gross total tumor resection, and had received adjuvant treatments (radio- and chemotherapies). None of these patients had had surgery for recurrent tumor resection and none had harbored multifocal tumors. The median age was 50 years. The median Karnofsky performance score was 80. Multivariate analysis using the Cox regression model revealed that the strongest prognostic variable was the amount of tumor necrosis on preoperative scan (P < 0.001), with median survivals of 42, 24, 15, and 12 months for tumor necrosis grades of 0 (7 'pts'), I (11 'pts'), II (9 'pts'), and III (21 'pts'), respectively. The intensity of enhancement of the tumor nodule was another prognostic factor (P = 0.003), with median survivals of 35, 18, and 13.5 months for enhancement grades of 0 (2 'pts'), I (22 'pts'), and II (24 'pts'), respectively. The extent of peritumoral edema had a quadratic effect (P = 0.001), with grades I (19 'pts'), II (22 'pts'), and III (7 'pts') surviving for 24, 12, and 20 months respectively. Location and volume of tumors were not statistically significant predictors of survival (P < 0.05). In conclusion, in this highly selected group, GBM patients with little or no necrosis and with less tumor nodule enhancement on preoperative MRI survive longer than patients with greater amounts of necrosis and greater degrees of tumor enhancement. In addition, a moderate degree of peritumoral edema is associated with worse prognosis.

Intracranial metastatic melanoma: correlation between MR imaging characteristics and melanin content.

OBJECTIVE: Preliminary reports based on limited numbers of cases have proposed that specific MR imaging patterns may permit a distinction between melanotic and amelanotic brain metastases in melanoma patients. The purpose of this study was to test this hypothesis by categorizing MR images obtained from a large series of patients and correlating the results with the percentage of melanin-containing cells in surgically resected metastases. SUBJECTS AND METHODS: The MR images of 30 patients with histologically proven intracerebral melanoma were reviewed retrospectively. Precontrast MR images were obtained with T1-weighted spin-echo sequences in axial and sagittal sections and with proton density-weighted and T2-weighted sequences in axial sections. After IV injection of gadopentetate dimeglumine (0.1 mmol/kg of body weight), T1-weighted images were obtained in axial and coronal sections. All patients had undergone gross total resection of the evaluated lesions. MR images of the metastases were reviewed and sorted into four groups on the basis of putative patterns: (1) melanotic pattern--hyperintense in relation to cortex on T1-weighted images, hypointense in relation to cortex on T2-weighted images, and isointense or hyperintense in relation to cortex on proton density-weighted images; (2) amelanotic pattern--hypointense or isointense in relation to cortex on T1-weighted images and hyperintense or isointense in relation to cortex on T2-weighted and proton density-weighted images; (3) indeterminate, or mixed, pattern--MR imaging characteristics that did not conform to those of one of the first two categories; and (4) hematoma pattern--MR imaging features that exhibited only hematoma characteristics. Tissue sections from all evaluated lesions were independently reviewed by a neuropathologist (G.N.F.), and the percentage of melanin-containing tumor cells in each resected metastatic lesion was estimated. The MR imaging data and histologic data were then compared to assess the predictive value of the MR imaging patterns. RESULTS: Forty-two metastatic lesions were identified and categorized by MR imaging pattern as follows: 10 melanotic, 11 indeterminate (mixed), 16 amelanotic, and five hematoma. Correlation with histologic findings revealed that a majority (7/10) of lesions that exhibited a melanotic MR imaging pattern had more than 10% melanin-containing cells, over half (9/16) of lesions that exhibited an amelanotic MR imaging pattern contained histologically identifiable melanin (but always in less than 10% of cells), and lesions that exhibited a mixed MR imaging pattern were either amelanotic or contained less than 10% melanotic cells. Conversely, a majority of lesions containing more than 10% melanotic cells (7/8) demonstrated the typical melanotic MR imaging pattern, lesions with less than 10% melanin-containing cells exhibited a variety of MR imaging patterns, and only about half of patients with amelanotic lesions (6/13) showed the characteristic amelanotic MR imaging pattern. For five lesions, potentially informative imaging data on melanin content was obscured by histologically documented hematoma formation. CONCLUSION: Only a minority of melanoma metastases have the anticipated MR imaging findings of melanotic melanoma, which consist of high signal intensity relative to that of cortex on T1-weighted images and low signal intensity relative to that of cortex on T2-weighted images. Of tumors that do exhibit this melanotic pattern, the majority have more than 10% melanin-containing cells. The putative MR imaging pattern for amelanotic melanoma is nonspecific, as over half of tumors with this pattern contain melanin.