A quantitative study of the EMI values obtained for normal brain cerebral infarction and certain tumours

A quantitative study has been made of the EMI numbers of normal brain, cerebral infarction and certain tumours. The scans were recorded on magnetic tape and analysed using a minicomuter linked to a graphic display unit. This system not only yielded 16 grey scales compared with the ten currently available, but was programmed to allow selected regions of the scans to be outlined. From these regions the computer calculated the area, the mean EMI number and its standard deviation. It was found that in 15 normal brain scans, the EMI values obtained for normal frontal and temporal lobes were similar, but that the values for the basal ganglia and occipital lobes were significantly different from the first two regions and from each other. Ten cases of cerebral infarction and 30 cases of cerebral tumour were analysed, and it was shown that analysing representative areas was more informative than surveying the whole lesion. Whilst only half of the scans of brain tumours had a significantly altered EMI number compared with that of normal brain, enhancement of tumour density with sodium iothalamate revealed a consistent and significant elevation of the EMI number for all tumours. In particular, the value for enhanced meningiomas was almost double and malignant tumours more than a third larger than normal brain. It was not possible to differentiate quantitatively between astrocytomas and metastases.     

Altered expression of the p53-regulated proteins, p21Waf1/Cip1, MDM2, and Bax in ultraviolet-irradiated human skin

The gemistocytic astrocytoma is a histological variant of diffuse astrocytomas and is characterised by the presence of large, GFAP-expressing neoplastic astrocytes (gemistocytes) and a tendency towards rapid progression to glioblastoma. In this study, we analyzed 28 gemistocytic astrocytomas (mean fraction of gemistocytes, 35.0+/-9.9%) for mutations in the p53 and PTEN (MMAC1) tumour suppressor genes. Single strand conformation polymorphism (SSCP), followed by direct DNA sequencing of p53 exons 5-8, revealed a mutation in 23 of 28 (82%) cases. Regional analysis of four tumours revealed identical p53 mutations in gemistocytic and fibrillary tumour areas. In contrast, none of 15 gemistocytic astrocytomas (WHO Grade II) and only two of 11 (18%) anaplastic gemistocytic astrocytomas (WHO Grade III) contained a PTEN mutation. Of these, one was a 1 bp deletion in codon 345 and the other a 1 bp insertion in intron 4. Differential PCR did not reveal homozygous PTEN deletion in any of the tumours analysed. These results indicate that p53 mutations are a genetic hallmark of gemistocytic astrocytomas, whilst PTEN mutations are absent in low-grade and rare in anaplastic gemistocytic astrocytomas.     

Effects of nitric oxide modulation on tumour blood flow and microvascular permeability in C6 glioma

C6 glioma strongly express nitric oxide synthase. Rats bearing C6 tumours were pre-treated with i.v. Ng-nitro-L-arginine methyl ester (L-NAME), 3-morpholinosydnonimine (SIN-1) or saline before local cerebral blood flow (LCBF) or tumour capillary permeability (TCP) was measured by the [14C]iodoantipyrine autoradiographic or [14C]alpha-amino-isobutyric acid techniques. L-NAME and SIN-1 caused significant TBF alterations (-44% and +136%, respectively) with less marked (-15% and +33%) alterations in normal brain. Calculated cerebrovascular resistance changes within tumour were indeed selective. Baseline TCP was increased compared with normal brain (20-fold). L-NAME and SIN-1 administration did not alter TCP. These effects have significant implications for human malignant glioma management. Selective i.v. manipulation of LCBF, without significant changes in TCP, could increase the efficacy of chemotherapy, radiotherapy or provide better peritumoural oedema control.     

Effect of negative middle-ear pressure on transient-evoked otoacoustic emissions

OBJECTIVE: To seek an optimal treatment plan from the results of treatment for metastatic disease of the spine in children. DESIGN: An 8-year retrospective study of children with metastatic disease of the spine. Imaging studies were reviewed and treatment modalities analysed. SETTING: The divisions of pediatric orthopedics and pediatric neurosurgery at the Children's Hospital of Eastern Ontario, Ottawa. PATIENTS: All children seen between April 1980 and December 1987 who had lesions metastatic to the spine by hematogenous or direct extension. There were 20 children (15 boys, 5 girls) with a mean age at the time of diagnosis of 9.5 years. Follow-up ranged from 2 weeks to 108 months. One child was lost to follow-up. INTERVENTIONS: Eleven children underwent laminectomy and decompression. Of the 14 neurologically compromised children, 5 received chemotherapy and radiotherapy and 9 received chemotherapy, radiotherapy and surgery. MAIN OUTCOME MEASURES: Type of metastatic lesion, vertebrae involved and response to therapy. RESULTS: Vertebrae involved with metastases were as follows: cervical (3), thoracic (5), lumbar (8) and multilevel (2). Meninges were involved in 2 cases. The most common causes of metastatic spinal involvement were neuroblastoma (4 cases) and astrocytoma (6 cases). Pathologic fractures occurred in 4 children and kyphoscoliosis in 4. Spinal cord paresis developed in 14 of the 20 children. Of the 6 children who survived from 48 to 108 months, 5 had tumours of neural origin, 4 being astrocytomas. Children with neuroblastoma or leukemic infiltration had a good initial response to chemotherapy. Five of the 6 surviving children had astrocytomas, and 5 were treated by surgical decompression. CONCLUSIONS: Metastatic disease of the spine in children secondary to astrocytoma should be treated aggressively, but from the experience gained from this study it is impossible to devise a rigid treatment plan for each type of metastatic tumour. The choice of chemotherapy, radiotherapy or surgery depends on the type of tumour, the age of the child and whether or not the spinal cord is compromised.     

Production of urokinase-type plasminogen activator (u-PA) and plasminogen activator inhibitor-1 (PAI-1) in human brain tumours

We investigated the role of plasminogen activators (PAs) and their inhibitor (plasminogen activator inhibitor-1, PAI-1) in human brain tumours. The amounts of urokinase-type plasminogen activator (u-PA), tissue-type plasminogen activator (t-PA), and plasminogen activator inhibitor-1 (PAI-1), and the activity of u-PA and t-PA were determined by enzyme-linked immunosorbent assay (ELISA), and u-PA and PAI-1 were immunolocalized using monoclonal antibodies in human brain tumours and normal brain tissues. The tissues were surgically removed from 64 patients; normal brain tissue (5 cases), low-grade glioma (4 cases), high-grade glioma (17 cases), metastatic tumour (9 cases), meningioma (benign 12 cases, malignant 6 cases), acoustic schwannoma (11 cases). u-PA activity and u-PA and PAI-1 antigen levels were significantly elevated in malignant brain tumours (malignant meningiomas, high-grade gliomas, and metastatic tumours) and acoustic schwannomas but very low in benign meningiomas, low-grade gliomas and normal brain. There was no difference in t-PA antigen levels among normal and malignant tissues, however levels of t-PA activity were markedly decreased in metastastic tumours. All malignant brain tumour tissues showed positive immunostaining for u-PA and PAI-1, however, some tumour cells showed negative intensity while others showed strong intensity for these antibodies. This contrasts to the homogeneous staining pattern found in acoustic schwannoma. These findings indicate that malignancy in human brain tumours is associated with elevated levels of u-PA and PAI-1 and that an imbalance between these proteins in a micro-environment contributes (ascribes) to tumour cell invasion.     

Cellular targets of exogenous tumour necrosis factor-alpha (TNF alpha) in human gliomas

To identify the cellular targets of TNF alpha in human gliomas, a total of 30 surgical specimens (12 glioblastomas, 4 anaplastic astrocytomas, 3 astrocytomas, 7 brains adjacent to tumour (BAT), 4 histologically normal-appearing brains) were examined by in vitro binding technique using biotinylated TNF alpha. The TNF-binding sites (TNF-BS) were recognized in the tumour cells in 8 of the 12 glioblastomas, 3 of the 4 anaplastic astrocytomas and in all the 3 astrocytomas. The TNF-BS were also recognized in the vascular endothelial cells in all these cases. The presence of TNF-BS in blood vessels ranged from 7.7 to 74.4% of the background vessels. This wide range of variation in the presence of TNF-BS within the tumour cells and tumour blood vessels may be relevant to the variable response of individual tumours to TNF alpha therapy. Since the tissue of normal brain, which lacks TNF-BS, might hardly be affected by this cytokine, administration of TNF alpha may be considered as an adjuvant therapy in selected groups of patients.     

Loss of inhibitory synapses on the soma and axon initial segment of pyramidal cells in human epileptic peritumoural neocortex: implications for epilepsy

The peritumoural neocortex removed from epileptic patients represents an important region for research because of its possible relationship to the generation, maintenance, and propagation of seizures. The peritumoural neocortex removed from an epileptic patient showing a regrowth of an anaplastic astrocytoma was examined in detail using immunocytochemistry for gamma-aminobutyric acid, glutamic acid decarboxylase, parvalbumin, nonphosphorylated neurofilament protein, glial fibrillary acidic protein, and histocompatibility antigen HLA-DR. The patterns of immunostaining were compared with the cytoarchitecture and myeloarchitecture in adjacent sections, and with the patterns of immunostaining observed in normal control neocortex. Furthermore, quantitative electron microscopy was used to compare the synaptic densities of presumptive excitatory and inhibitory synapses between regions showing different grades of cytoarchitectural and neurochemical alterations in the peritumoural neocortex, and to compare these regions with normal neocortex. A variety of changes in synaptic circuits in the peritumoural neocortex was found, but it appears that neurons within the less abnormal-looking regions were involved in altered synaptic circuits that might contribute to epileptic activity. In these regions, the most prominent change was the loss of inhibitory synapses on the soma and axon initial segment of pyramidal cells, but numerous excitatory synapses were present on their dendrites that would make these neurons hyperexcitable. However, the most abnormal regions histologically were likely a primary zone for progression of the tumour, with many surviving neurones, but which received and formed very few synapses; thus, they were probably unrelated to the initiation, maintenance, or propagation of seizures.     

Delta-1 is a regulator of neurogenesis in the vertebrate retina

Controversy exists about the effect of tissue edema on cerebral microcirculation. High spatial resolution is required for observation of extravasation and microcirculation during focal vasogenic edema formation. To study the relationship between tissue edema and perfusion, we developed a technique for simultaneous visualization of extravasation and microvessel perfusion in rats. Focal intracortical microvascular injury was generated with a 1-sec Nd-YAG laser pulse. Evans blue albumin (EBA) was infused 30 min before decapitation to study extravasation and FITC-dextran was injected 30 sec prior to decapitation to examine microvessel perfusion. Computerized scanning laser-excited fluorescence microscopy followed by high resolution image analysis permitted quantitative assessment of both parameters on single fresh-frozen brain sections. Studied at 30 min (3.66 +/- 0.15 mm), 2 hr (4.14 +/- 0.08 mm, P < .05), and 8 hr (4.69 +/- 0.18 mm, P < .01) after injury, the diameter of the circular, sharply demarcated zone of EBA-extravasation increased progressively. At 30 min, microvessels at a zone surrounding the area of EBA-extravasation contained 69 +/- 14% (P < .05) more fluorescent FITC-filling than in the control hemisphere, but the density of perfused microvessels was unchanged. At 2 hr, secondary tissue changes had already occurred in a zone surrounding the initial laser lesion. While severe reduction in the density (-76 +/- 13%, P < .05) of perfused microvessels was observed within 400 to 240 microm inside the border of EBA extravasation, perfusion indexes were normal despite the presence of extravasated plasma constituents within 0-80 microm from the border. In a narrow zone (80 microm) outside the border of extravasation, individual microvessels contained 34 +/- 9% (P < .01) less FITC-fluorescence than those in a homologous area of the uninjured contralateral hemisphere. This report demonstrates the feasibility of simultaneous measurement and high-resolution mapping of indices of microvascular perfusion (density, filling) and extravasated plasma constituents in damaged and intact brain areas. In this model, the presence of extravasated plasma constituents the size of proteins did not immediately influence indices of cortical microcirculation. However, microvascular perfusion may be perturbed surrounding such an area of advancing vasogenic edema formation.     

Genetic basis of neurological tumours

Neurological tumours are common neoplasms of both adults and children. Recent studies have begun to delineate the genetic abnormalities that underlie such tumours, and have implicated two classes of genes, oncogenes and tumour suppressor genes. Most investigations have focused on those astrocytomas that affect the cerebral hemispheres of adults, since these are the most common and malignant brain tumours. The high-grade astrocytomas that affect adults, such as glioblastoma multiforme, often have amplification of the epidermal growth factor receptor (EGFR) oncogene and loss of a variety of chromosomal loci that probably harbour tumour suppressor genes. Of the various tumour suppressor gene loci, the p53 gene on chromosome 17p has been studied most closely and has been shown to be mutated in both low- and high-grade astrocytomas. These genetic alterations may provide a means for subdividing astrocytomas into diagnostic categories. For instance, p53 gene mutations occur more commonly in glioblastomas from young adults and women, while EGFR gene amplification is more common in glioblastomas from older adults and men. For the other primary CNS tumours, genetic studies remain in their infancy. The neurocutaneous syndromes, such as neurofibromatosis types 1 and 2, have provided unique insights into neurological oncogenesis. The NF1 gene on chromosomes 17q and its product, neurofibromin, may be important in the formation of neurofibrosarcomas, while the NF2 gene on chromosome 22q and its product, merlin, are probably involved in the formation of schwannomas and other nervous system tumours. The further characterization of these and other neurological tumour genes will undoubtedly illuminate many other areas in neurooncology.     

Distribution of endogenous tumour necrosis factor alpha in gliomas

AIMS: To determine the distribution and cellular origin of endogenous tumour necrosis factor alpha (TNF alpha) in the cellular components of human gliomas. METHODS: Frozen sections of 26 gliomas (four astrocytomas (As); two oligoastrocytomas (OA); one ansplastic astrocytoma (AA); one anaplastic oligoastrocytoma (AOA); 18 glioblastomas (GB)) were examined immunohistochemically using antihuman TNF alpha and anti-Leu-M5 (CD11c) antibodies. Additional studies with double immunohistocchemical procedures were performed with anti-glial fibrillary acidic protein and anti-neurofilament antibodies. RESULTS: Eighty per cent of the AA, AOA, and GB (16 of 20) had a positive reaction for TNF alpha, but only 17% of As and OA (one of six) were positive. Positive cells were seen in both the tumour tissue and adjacent brain tissues. TNF alpha protein was detected not only in the tumour cells but also in the endothelium of tumour vessels as well as reactive astrocytes and neurons. CONCLUSIONS: Endogenous TNF alpha is present in cells of various origins in glial tumours including tumour vessels; however, the role of TNF alpha may be different in different types of cells or altered microenvironment.     

Epidemiology and prognosis in children treated for intracranial tumours in Denmark 1960-1984

A total of 911 Danish children under 15 years of age were treated for an intracranial tumour in the 25-five year period 1960-1984. All cases were followed up to the end of 1994 or to emigration or death if one of these came sooner. The mean annual incidence was 32.5 per million children with a slight increase over the 25 years. The male/female ratio was 1.15 and close to the M/F ratio for the entire Danish population of children. Of the tumours, 46% were located in the supratentorial and 54% in the infratentorial compartment, and 94% were verified histologically. In order of frequency the most common types were astrocytomas (all grades, 35%), medulloblastomas (20%), ependymomas (14%), and craniopharyngiomas (5%). Total removal of the tumour was performed in 277 and partial removal, including biopsy, in 490 children. In 57 patients a shunt operation only was performed, and 87 children did not have an operation or died before the correct diagnosis was established. Radiotherapy was administered in 55%. The outcome depended on extent of removal, radiation, location and histology of the tumour. Most (784 or 86%) of the children survived more than 1 month after diagnosis or operation, and 353 children (39% of the whole series, 47% of those alive more than 1 month after diagnosis) were alive at follow-up. Of the survivors 29% had a tumour in the supratentorial midline, 26% one in the lateral part of the supratentorial area, 31% a cerebellar tumour and 13% a IV ventricle tumour. It was possible for 66% of the survivors with supratentorial and 90% of those with infratentorial tumours to lead a normal life. The long-term prognosis was especially good for children with cerebellar and supratentorial astrocytomas and optic chiasma tumours. Children with juvenile cerebellar astrocytoma had the best prognosis: 90% were alive at the end of the follow-up period, as against 20% of those with medulloblastoma and 6% of those with glioblastoma. A comparison of the data from the present series and from a similar Danish series of intracranial tumours in 533 children seen in the years 1935-1959 shows no significant differences in location or histology, a slight increase in annual incidence, and improved survival rates during the 50 years in question.     

5-Aminolevulinic acid induced endogenous porphyrin fluorescence in 9L and C6 brain tumours and in the normal rat brain

A new approach in photodynamic therapy is the use of endogenous porphyrins for sensitisation of tumours to light. The induction of endogenous porphyrins after intravenous injection of 5-aminolevulinic acid (ALA, 200 mg kg-1) was studied in 23 rats, bearing intracranial 9L or C6 tumours. After 0, 2, 4, 6, 8, and 22 hours the rats were sacrificed and the fluorescence distribution of endogenous porphyrins was studied in brain tissue sections with a standard fluorescence microscope and a confocal laser scanning microscope. The role of blood-brain barrier disruption on porphyrin production was studied in 2 rats with a cryo-lesion of the cortex. Additionally, 9L and C6 tumour cell cultures were incubated with ALA for 8 hours in vitro. Fluorescence was measured with a fluorescence spectrophotometer in cell cultures and in the brain sections. Porphyrins were detected in vitro in the tumour cells from 2 hours onwards and ex vivo in the tumour sections mainly from 2 to 8 hours, by 22 hours porphyrin fluorescence had almost disappeared. The contralateral brain showed low fluorescence levels between 2 and 6 hours after ALA administration. At the site of the cryo-lesions low fluorescence was measured 6 hours after ALA administration. The 9L tumours fluoresced homogeneously, with a sharp demarcation towards normal brain tissue. Fluorescence in the C6 tumours was patchy, with a poorly fluorescing edge. In both tumour models fluorescence was also detected in brain surrounding the tumour and sometimes in contralateral white matter and ventricle ependyma and pia mater. The slight increase of porphyrin fluorescence in the normal brain of tumour bearing rats, compared to the absence of this in rats without a tumour, was attributed to transport by bulk flow of porphyrins made in the tumours, and possibly also of circulating porphyrins or ALA leaking from the tumour vessels.     

The role of PET-FDG in questionable diagnosis of relapse in the presence of radionecrosis of brain tumors

INTRODUCTION: Although CT and MR are sensitive techniques for the detection of cerebral tumours, both have limitations in distinguishing between tumour relapse (TR) and post-treatment radionecrosis (RN). PATIENTS AND METHODS: In this study we have determined the usefulness of metabolic imaging with PET-FDG in such situations. We assessed 70 patients with CNS tumours (22 low grade astrocytomas, 25 high grade astrocytomas, 3 oligodendrogliomas, 13 metastatic tumours and 7 other tumours. All had been treated with radiotherapy and other treatments such as radiosurgery, chemotherapy or different types of surgery, and presented clinical pictures which made it necessary to decide the differential diagnosis of relapse or radionecrosis. RESULTS: In the PET-FDG study visual and semiquantitative analysis was done by SUV (Standardized Update Value). Confirmation of the findings was obtained in 44 cases (24 TR and 20 RN). MR was doubtful or inconclusive in most cases, whilst with PET correct diagnosis was made in all cases. CONCLUSIONS: Metabolic imaging with PET-FGD is better than anatomostructural imaging techniques for differential diagnosis between tumour relapse and radionecrosis in CNS tumours which have been treated. Prospective studies are necessary for evaluation of SUV as a factor for prognosis of survival.     

Treatment of peptic ulcer

AIMS: To establish whether MIB-1 and p53 staining are useful for differentiating pilocytic astrocytomas and astrocytomas from anaplastic astrocytomas and glioblastomas. This study was restricted to children and young adults under 30 years of age because of the differences in p53 mutations between paediatric and adult astrocytomas. METHODS AND RESULTS: Forty-five astrocytic tumours, including 18 pilocytic astrocytomas, 14 astrocytomas, four anaplastic astrocytomas and nine glioblastomas, from 45 children and young adults, between 1 and 29 years (mean 11 years) of age, were examined pathologically, and sections from paraffin-embedded blocks were used for MIB-1 and p53 immunostaining. The MIB-1 labelling index and the frequency and intensity of p53 staining in both the pilocytic astrocytoma and the astrocytoma group were significantly lower than in the anaplastic astrocytoma plus glioblastoma group (P < 0.001). In 11.1% (two of 18) of pilocytic astrocytomas and 42.9% (six of 14) of astrocytomas, immunoreactivity of either MIB-1 or p53 staining was of almost the same intensity as that of anaplastic astrocytomas and glioblastomas. However, in these cases, results using both MIB-1 and p53 stain differed from those for anaplastic astrocytomas and glioblastomas. CONCLUSIONS: MIB-1 and p53 co-staining is very useful for differentiating pilocytic astrocytomas and astrocytomas from anaplastic astrocytomas and glioblastomas. However, MIB-1 or p53 staining alone cannot differentiate pilocytic astrocytomas and astrocytomas from anaplastic astrocytomas and glioblastomas.     

Prenatal diagnosis of a highly undifferentiated brain tumour--a case report and review of the literature

Intracranial tumours, often presenting with progressive hydrocephalus, are rare congenital diseases accounting for 0.5-1.5 per cent of all cases of brain tumours diagnosed during childhood. The differential diagnosis includes vascular malformations, infarctions, and haemorrhages. Sonographic signs suggestive of glioblastoma, teratoma, and astrocytoma do not establish the histological diagnosis, however. We report a case of an undifferentiated fetal glioma detected at 29 weeks' gestation. The diagnosis of an undifferentiated brain tumour was suspected by sonography because of the lack of normal brain structures in conjunction with a diffuse echogenic central lesion and an external hydrocephalus. Because of the very poor prognosis, we induced labour by intravaginal and intravenous administration of prostaglandin E2 and achieved the vaginal delivery of a stillborn child whose head circumference corresponded to 38 weeks of pregnancy. Histological and immunochemical features of this undifferentiated congenital glioma (glioblastoma) are presented.     

Clinical studies of oral antiemetic drugs on delayed emesis induced by cancer chemotherapy]

BACKGROUND AND PURPOSE: Our purpose was to develop a classification scheme and method of presentation of in vivo single-voxel proton spectroscopic data from astrocytomas that most closely match the classification scheme determined from biopsy specimens. Since in vivo proton spectroscopy is noninvasive, it may be an attractive alternative to intracranial biopsy. METHODS: Single-voxel spectra were acquired using the point-resolved spectroscopic pulse sequence as part of the Probe spectroscopy package on a G.E. 1.5-T Signa scanner. Subjects consisted of 27 patients with biopsy-confirmed brain tumors (13 with glioblastoma multiforme, six with anaplastic astrocytoma, and eight with low-grade astrocytoma). The patients were divided into groups based on the histologic subtype of their tumor for different treatment protocols. RESULTS: Metabolic peak areas were normalized for each metabolite (choline, creatine, N-acetylaspartate, lactate) to the area of the unsuppressed water peak and to the area of the creatine peak. Kruskal-Wallis nonparametric analysis of variance (ANOVA) tests showed statistically significant differences among the tumor groups for all the area ratios. The lactate/water ratio could be used to distinguished all three tumor groups, whereas the choline/water ratio distinguished low-grade astrocytomas from the two high-grade groups. Both the choline and lactate ratios could be used to separate the high-grade from the low-grade tumors. CONCLUSION: Specific relative metabolic peak area ratios acquired from regions of contrast-enhancing brain tumor can be used to classify astrocytomas as to histopathologic grade.     

Towards a method for automated classification of 1H MRS spectra from brain tumours

Recent studies have shown that MRS can substantially improve the non-invasive categorization of human brain tumours. However, in order for MRS to be used routinely by clinicians, it will be necessary to develop reliable automated classification methods that can be fully validated. This paper is in two parts: the first part reviews the progress that has been made towards this goal, together with the problems that are involved in the design of automated methods to process and classify the spectra. The second part describes the development of a simple prototype system for classifying 1H single voxel spectra, obtained at an echo time (TE) of 135 ms, of the four most common types of brain tumour (meningioma (MM), astrocytic (AST), oligodendroglioma (OD) and metastasis (ME)) and cysts. This system was developed in two stages: firstly, an initial database of spectra was used to develop a prototype classifier, based on a linear discriminant analysis (LDA) of selected data points. Secondly, this classifier was tested on an independent test set of 15 newly acquired spectra, and the system was refined on the basis of these results. The system correctly classified all the non-astrocytic tumours. However, the results for the the astrocytic group were poorer (between 55 and 100%, depending on the binary comparison). Approximately 50% of high grade astrocytoma (glioblastoma) spectra in our data base showed very little lipid signal, which may account for the poorer results for this class. Consequently, for the refined system, the astrocytomas were subdivided into two subgroups for comparison against other tumour classes: those with high lipid content and those without.     

Histologically confirmed changes on CT of reoperated low-grade astrocytomas

During a 15-year period 37 patients with primary low-grade astrocytoma have been operated upon twice in our institute. CT and histological data at the time of the first and second operations were compared. The majority of primary astrocytomas showed as a low-density area without contrast enhancement; minimal, mainly marginal enhancement was present in six cases. At reoperation 10 tumours were histologically unchanged; the corresponding CT studies displayed a nonenhancing lesion, although insignificant contrast uptake could be seen in three cases. There were 18 tumours which had transformed into anaplastic astrocytoma: CT before repeat surgery showed more or less marked enhancement. In all nine cases which progressed into glioblastoma multiforme strong contrast enhancement was seen on CT at the time of recurrence. Although the grade of contrast uptake varied, the appearance of or increase in enhancement was a sign of some degree of anaplastic change, particularly convincing in cases of dedifferentiated glioblastoma multiforme.