Acta Neurochir (Wien) (2004) 146: 19–26 DOI 10.1007/s00701-003-0151-x
Clinical Article Distinct supratentorial lesions mimicking cerebral gliomas G. Wurm1 , B. Parsaei1 , R. Silye2 , and F. A. Fellner3;4 1
Department of Neurosurgery, Landesnervenklinik Linz, Linz, Austria Department of Pathology, Landesnervenklinik Linz, Linz, Austria 3 Institute of Neuroradiology, Landesnervenklinik Linz, Linz, Austria 4 Institute of Diagnostic Radiology, Friedrich-Alexander University, Erlangen-N€urnberg, Germany 2
Published online December 9, 2003 # Springer-Verlag 2003
Summary Background. Various supratentorial pathological conditions can mimic neoplastic cerebral lesions clinically as well as radiologically. Analysis of the neuroradiological findings, the clinical history, laboratory and other paraclinical data mostly help to narrow down the diagnosis of cerebral pathologies. Sometimes, however, histopathological analysis of the operative specimen after surgery reveals unexpected findings. Patients and findings. In a series of 197 operative procedures performed for glioma surgery between August 2000 and August 2002 we found six distinct cases mimicking supratentorial tumours. Clinicoradiological findings had suggested a neoplastic gliomatous process in all cases. But histopathological examination revealed that in reality one patient had been affected by a stroke, two by hypertensive encephalopathy, and one by radiation necrosis; multiple sclerosis was the underlying pathology in two patients. Interpretation. Conditions which show features similar to those of neoplastic cerebral lesions require advanced magnetic resonance imaging (MRI). The benefit of such sophisticated imaging in relation to the costs is an important issue in this context. Further research in the field of modern image modalities is necessary to evaluate these noninvasive techniques for specification of intracerebral lesions. Keywords: Glioma; misdiagnosis; stroke; hypertensive encephalopathy; radiation necrosis; multiple sclerosis.
Introduction Distinguishing brain tumour from other lesions in patients presenting with acute focal neurological signs and symptoms is crucial to avoid improper treatment or delayed treatment. Such a differentiation based on clinical status and conventional structural imaging may however be difficult. Numerous tumoural diseases (namely lymphoma, metastasis, developmental tumours) and non-tumoural diseases (vascular, malformative,
iatrogenic and inflammatory conditions) may masquerade as glioma on conventional images. This diagnostic dilemma seems to be especially true for the following pathological conditions: Sometimes, the clinical symptoms caused by a stroke mimic those of a brain tumour or vice versa. In 1999, Morgenstern could show that even nowadays 4.9% of the patients discharged with the diagnosis of brain tumour were initially thought to have suffered from a stroke [16]. The percentage of stroke patients whose pathology mimicks that of brain tumour at first is unknown in the literature. Hypertensive encephalopathy is a complex cerebral disorder with inconclusive symptoms including headache, seizures, visual disturbances, altered mental status and focal neurological signs. It is associated with a variety of conditions in which blood pressure rises acutely and it has been ascribed to either exuberant vasospasm with ischemia or to vasodilatatory mechanisms with breakdown of autoregulation and interstitial edema [18, 19]. This pathology rarely mimicks the features of a gliomatous tumour [18]. Radiation necrosis is a typical complication after radiation therapy for extra- and intracranial tumours. Many of these patients present with only vague clinical signs [4, 5]. The frequent diagnostic dilemma of neoplasm versus radiation necrosis is addressed in many studies [3, 6, 11]. Multiple sclerosis may in a few cases present as a cerebral mass with the clinical features and neuroradiological appearance of a cerebral tumour [1, 2, 8, 20].
20
Usually, the typical radiological feature of multiple sclerosis is the lack of mass effect as well as of edema even in large white matter lesions. Nevertheless, plaque lesions with mass effect, a poor clinical course, rapid deterioriation and=or edema cause major diagnostic challenges as they are mostly believed to be characteristics of a tumour [1, 8, 24]. Patients and findings In a period of two years (August 2000 to August 2002), 197 operative procedures have been performed for glioma surgery in the neurosurgical department in Linz, Austria. Within this period, six cerebral lesions were falsely interpreted as gliomatous tumours preoperatively by means of conventional MR imaging (MRI); four of the six MRIs had been performed in foreign hospitals. Conventional MRI included T2-weighted turbo spin-echo, fluid-attenuated inversion recovery (FLAIR), T1weighted spin-echo before, and T1-weighted spin-echo (with magnetization transfer pulse) after i.v. application of Gd-DTPA. This investigation battery showed that each of these six lesions were accompanied by cranial mass effect and other features typical for cerebral gliomas. Eventually, one lesion was diagnosed as a stroke, two as hypertensive encephalopathy, one as radiation necrosis; in two cases multiple sclerosis was found out to be the underlying pathology. The resected tissue was completely assessed by histopathology, thus excluding a sampling error. A second opinion was requested from other neuropathology laboratories in each of the cases, which confirmed our histopathological diagnoses. None of the six patients would have been operated on, if the true nature of the lesion had been
G. Wurm et al. recognized beforehand. Four illustrative cases will now be presented in order to warn of possible misdiagnoses of such lesions:
Case 1 A left-handed 57-year-old male was admitted to our clinic complaining of having suffered from headache and focal seizures for three months. Neurological examination on admission revealed no speech, sensory or motor disturbances. Conventional MRI including the above mentioned investigations were performed showing a discrete inhomogeneous mass in the right frontolateral lobe without significant contrastenhancement but with slight perifocal edema; the lesion was interpreted as low-grade astrocytoma or dysembryoblastic neuro-epithelial tumour (DNET) (Fig. 1, left). Thus, a frontolateral craniotomy was performed; the speech area was delineated by means of cortical stimulation via subdural elecrodes; resection of the presumptive tumour was guided by neuronavigation. Histological examination of the resected specimen (seven paraffine blocks, no unexamined tissue left) surprisingly showed cerebral infarction with necrosis and macrophage invasion; the surrounding brain tissue was reactive also in Ki67 immunohistochemistry; there was no evidence of a neoplasma (Fig. 1, right). The patient was discharged five days after operation without any neurological disturbances. A follow-up MRI 3 months later showed no residue of the lesion in the resection area.
Case 2 A 76-year-old female was admitted with a 20 year history of chronic hypertension. Clinical symptoms for two weeks included headache, blurred vision and impaired cognitive functions. Gadolinium enhanced
Fig. 1 (Case 1). Left: Conventional MR imaging (A T2-weighted turbo spin-echo, B FLAIR, C T1-weighted spin-echo before, and D T1-weighted spin-echo (with magnetization transfer pulse) after i.v. injection of Gd-DTPA): Discrete inhomogeneous mass in the right frontolateral lobe (frontal operculum) without significant contrast-enhancement. Slight perifocal edema suspicious of low-grade astrocytoma or DNET. Right: Cerebral infarction with necrosis and macrophage invasion organising necrotic areas; no evidence of neoplasm
21
Distinct supratentorial lesions mimicking cerebral gliomas
Fig. 2 (Case 2). Left: Conventional MR imaging (A T2-weighted turbo spin-echo, B FLAIR, C T1-weighted spin-echo before, and D T1-weighted spin-echo (with magnetization transfer pulse) after i.v. injection of Gd-DTPA): Gadolinium enhanced MR demonstrates a focally contrast enhancing mass in the right temporal pole suspicious of low grade glioma with focal malignization. Right: Hypertensive encephalopathy with hyalinisation of small vessels and infarction within white matter MR demonstrated a focally contrast enhancing mass in the right temporal pole suspicious of low-grade glioma with focal ‘‘malignization’’ (Fig. 2, left). The assumed tumour was totally removed via temporal craniotomy. Histological examination then showed typical signs of hypertensive encephalopathy: hyperplasia of the peripheral vessel wall was combined with obliteration and recanalisation; the brain parenchyma showed fibrinoid degeneration and infarctions. Five paraffineblocks were taken from the 4:3:1 cm specimen, no unexamined tissue was left. Moreover, perivascular areas demonstrated haemorraghes with migration of macrophages and leucocyts, edema within the white matter as well as hyalinisation of the vessel walls, perivascular necrosis and glia proliferation, which was negative for Ki67 immunohistochemistry staining (Fig. 2, right). The postoperative course was uneventful and the symptoms disappeared after therapy for hypertension. A similar case (Case 3) is presented in Table 1.
Case 4 A 62-year-old man had a history of right tonsillar carcinoma treated with radiation therapy and chemotherapy in 1998. One year later, he developed amaurosis as a late side effect of radiation therapy. In summer 2000, he presented with headache and loss of weight. MRI showed an inhomogeneously contrast enhancing mass lesion in the right temporal lobe with invasion of the amygdala. There was considerable perifocal edema. These findings were interpreted as a highly malignant tumour such as glioblastoma multiforme. (Fig. 3, left). Because of this assumption a right sided temporal craniotomy was performed and the large space occupying lesion was removed. Histological examination of resected specimen (three representative blocks), however, revealed features of radiation necrosis; no mitotic activity or signs of malignancy were found (Fig. 3, right); Ki67 immunohistochemistry showed negative
Table 1. Clinicoradiological features of patients ID
Sex
Age
Symptoms
MR studies foreign=our
MR diagnosis
Location of lesion
Procedure operation=biopsy
Histopathology
1
m
57
seizures, headache
our
low-grade glioma
temporal
stroke
2
w
76
foreign
low-grade glioma
temporal
3
w
70
impaired vision and cognitive functions blurred vision, headache
image guided resection resection
foreign
low-grade glioma
occipital
4 5 6
m m w
62 33 52
headache, loss of weight hemiparesthesia hemiparesis, dizziness
our foreign foreign
high-grade glioma low-grade glioma low-grade glioma
temporal parietal parietal
image guided resection resection stereotactic biopsy image guided biopsy
hypertensive encephalopathy hypertensive encephalopathy radiation necrosis multiple sclerosis multiple sclerosis
22
G. Wurm et al.
Fig. 3 (Case 4). Left: Conventional MR imaging (A T2-weighted turbo spin-echo, B FLAIR, C T1-weighted spin-echo before, and D T1-weighted spin-echo (with magnetization transfer pulse) after i.v. injection of Gd-DTPA): Inhomogeneously contrast enhancing mass lesion in the right temporal lobe with invasion of the amygdala. There is considerable perifocal edema. These findings are typical of a highly malignant tumour, such as glioblastoma multiforme. Right: Radiation necrosis with signs of spongiosis, edema on the right side and proliferation of reactive capillaries and of glia on the left side of the figure; no mitotic activity or signs of malignancy were found
Fig. 4 (Case 5). Left: Conventional MR imaging (A T2-weighted turbo spin-echo, B FLAIR, C T1-weighted spin-echo before contrast injection, D Reformation from a 3D rapid gradient-echo data set (MP-RAGE) after gadolinium administration): On T2-weighted images a relatively homogeneous hyperintense lesion is found which demonstrates only moderate gadolinium enhancement on the T1-weighted spin-echo scan after contrast injection. In this case a low-grade glioma was suspected. Right: Neuronavigationally guided biopsy revealed a plaque of multiple sclerosis with loss of neuronal parenchyma, perivascular spaces and infiltration of lymphocytes and macrophages; no evidence of a tumourous process
23
Distinct supratentorial lesions mimicking cerebral gliomas staining pattern. The postoperative course was uneventful, general condition improved.
Case 5 A 33-year old female was admitted to our clinic having suffered from left sided hemiparesthesia for 6 weeks. On T2-weighted images a relatively homogeneous hyperintensive lesion was found which demonstrated only moderate gadolinium enhancement on the T1-weighted spin-echo scan after contrast application. In this case a low-grade glioma was presumed (Fig. 4, left). Neuronavigationally guided biopsy was performed and the histopathological specimen was discovered to be a plaque of multiple sclerosis; there was no evidence of a tumourous process (Fig. 4b). Steroid therapy helped to diminish the neurological disturbances. Similar features were found in Case 6 (Table 1).
Discussion Differential diagnosis of gliomas Today the diagnosis of a cerbral lesion is mainly based on various imaging methods, which may then be confirmed by clinical symptoms and signs and paraclinical findings. However, these methods may sometimes be insufficient, thus leading to surgical procedures that would not have been performed with a correct presurgical diagnosis. Due to similar clinical and neuroradiological features, special lesions are prone to be misdiagnosed as brain neoplasms. The spectrum of diseases potentially leading to a wrong diagnosis includes vascular, iatrogenic and inflammatory conditions: Stroke While CT is frequently employed for acute neurological deficits to exclude intracranial hemorrhage, it is not sufficient to prove or exclude brain tumour [16]. But even in patients with conventional MRI, vascular lesions cannot be distinguished from tumours in all of the cases. A glioblastoma for instance was falsley interpreted as an infarction, an infarction was misinterpreted as metastasis and a glioma was believed to represent a simple haemorrhage by mistake in the report of Jellema and collegues in 2001 [10]. In 1998, Okamoto reported of small cortical enhancing infarcts mimicking metastases in three patients [17]. The patient previously presented as case 1 of our report also belongs to this category of people whose pathologies were misinterpreted by MRI. Hypertensive encephalopathy The reversible leuco-encephalopathy syndrome, which is the neuroradiological correlate of hyperten-
sive encephalopathy always shows changes on T2weighted MR imaging. But its findings are mostly shown best on fluid-attenuated inversion recovery (FLAIR) MR images. Nevertheless, these conventional techniques alone are not sufficient to recognize all of these conditions, especially when lesions are located untypically and=or when presenting with mass effect and even focal enhancement [18]. Our patient number 2 shows that hypertensive encephalopathy can occur in unusual regions. An atypical distribution of this entity is occasionally seen in anterior locations (patient 2), the basal ganglia, cerebellum or brain stem. This case also confirms that morphological, clinical and histopathological phenomena are not always correlated [21]. Radionecrosis Radiotherapy is the primary treatment for nasopharyngeal carcinoma. Owing to the close proximity to the skull base, the medial temporal lobes are inevitably included in the target volume of irradiation. This fact seems to have adverse but only insignificant effects on the cognitive functions of the patients [5]. The reported incidence of radiation necrosis of temporal lobes varies from 0 to 18.6% depending on the dose regimen used [5, 7, 14]. Even if some authors recommend surgery for radiation necrosis [5, 13], the mild symptoms of our patient (Case 4) would not have been reason enough for a surgical procedure. If he had not been believed to harbour a glioblastoma, treatment with corticosteroids would have been the therapy of choice as it achieves durable objective responses in a considerable part of patients [12]. MS The neuroradiological evidence of a single, large white matter lesion with mass effect, clinically revealed by signs of endocranial hypertension, gives rise to the assumption of it being a central nervous system neoplasm. But in rare cases a demyelinating disorder can also start with the above mentioned characteristics. Moreover, the clinical history in patients with multiple sclerosis is not always conclusive [1, 2, 8]. Atypical imaging features led to the wrong diagnosis of brain tumours in our cases, number 5 and 6 for instance. Hence, a brain biopsy (one neuronavigationally guided and one stereotactically guided) was carried out, leading to an increase in time obligatorially spent in the hospital and to delay of correct treatment.
24
On could argue that the diagnoses of infarction and radiation necrosis should have been more strongly considered in Cases 1 and 4, respectively. Thus a ‘‘wait and see’’ policy in the first case and a steroid trial in the second could have avoided craniotomy. Additionally, stereotactic biopsy or PET scanning could have been adopted to obtain additional information in our cases. However, improper resection actually was performed due to misdiagnosis of conventional imaging. In this paper we wanted to show, first, that preoperative misdiagnosis is still a matter of fact, and second, that further research in the field of modern image modalities is warranted in order to evaluate these techniques for specification of intracerebral lesions. Advanced MR techniques Preliminary results suggest that advanced MR techniques such as diffusion-MR, perfusion-MR (dynamic susceptibility contrast MR) and MR-spectroscopy (MRS) may increase sensitivity and specificity for various cerebral lesions [7, 9, 15, 22, 23]. For instance, Sundgren and coworkers consider MRI diffusion and perfusion imaging to be pathognomonic for hypertensive encephalopathy; they could show that perfusion imaging reveals a considerable reduction in blood volume and flow with corresponding vasogenic edema in the affected region [21].
G. Wurm et al.
MR spectroscopy (MRS) that allows in vivo measurements of the relative concentration of brain metabolites may provide biochemical information helpful in distinguishing brain tumour from other pathologies [7, 15]. MR spectroscopy is used for better differentiation between tumour and radiation necrosis, based on the fact that choline is increased in tumours. However, in some cases of radiation necrosis, choline may be increased too. Therefore, differentiation between radiation necrosis and tumours may not be possible in a certain percentage of patients when using MRS alone. In contrast, relative regional cerebral blood volume (rrCBV) mapping through dynamic susceptibility contrast MRI (perfusion MR) seems to be more promising as reported by several authors [9, 22, 23]. Perfusion MR reveals marked hypoperfusion in radiation necrosis [22] in contrast to highly malignant brain tumours which show increased rrCBV when compared to normal brain tissue. There is legitimate hope that with more experience these advanced MR techniques can be used as a noninvasive method in presurgical evaluation [9]. These techniques are meanwhile commercially available and are no longer restricted to few specialized research centres. High technological standards and an individualised problem solving approach are prerequisites for this purpose. Additionally, a selective referral policy has to be developed and the cost=benefit ratio has to be clarified. Figure 5 shows an example of advanced MR imaging,
Fig. 5. Example of advanced MR imaging including perfusion-MR (dynamic susceptibility contrast MR) and MR-spectroscopy (MRS), where the assumption of astrocytoma grade II WHO was proven by means of tumour biopsy
25
Distinct supratentorial lesions mimicking cerebral gliomas
where the assumption of astrocytoma grade II WHO was proven by means of tumour biopsy (Fig. 5).
Cost=benefit evaluation The benefits in terms of better diagnosis of brain lesion offered by newer imaging techniques need to be considered along with the increase in cost. In the present climate of cost-containment, it is reasonable to ask if expensive diagnostic tools are always needed when conventional MRI provides sufficient information concerning diagnosis and etiology of a lesion in most cases. On the other hand, the high initial cost of presurgical evaluation for glioma surgery may be offset by avoidance of unnecessary surgical procedures and of postoperative hospital care. Additionally, the issue of delayed correct treatment and of adverse psychosocial effects for the patient have to be taken into consideration.
Further evaluation of modern imaging techniques The field of modern image modalities appears to herald an exciting future [7, 9, 15, 22, 23]. But further research in advanced MR imaging is warranted on a large scale in order to evaluate these noninvasive techniques for specification of intracerebral lesions. Evaluating the specifity of modern imaging techniques and assessing whether the increase in cost is justified by gains in preoperative visualization should thus be a short term aim. For this purpose, we started a prospective study on advanced MR techniques (diffusion studies, perfusion studies, spectroscopy) where we use imageguided or stereotactic biopsies from various pathologies obtained from distinct parts of the lesions.
Conclusion Despite tremendous advances in structural and functional neuro-imaging, distinguishing the presentation of brain tumour and other pathologies remains an important diagnostic challenge. Economic disadvantages as well as adverse psychosocial consequences associated with false preoperative assumptions have to be considered. Hence, conditions which show features similar to those of neoplastic cerebral lesions require advanced imaging investigations. Preliminary clinical data give rise to the hope that entirely noninvasive, specific diagnosis of brain tumours may become possible by means of enhanced MR imaging techniques.
References 1. Capello E, Roccatagliata L, Pagano F, Mancardi GL (2001) Tumor-like multiple sclerosis (MS) lesions: neuropathological clues. Neurol Sci 22 [Suppl]2: 113–116 2. Censori B, Agostinis C, Partziguian T, Gazzaniga G, Biroli F, Mamoli A (2001) Large demyelinating brain lesion mimicking a herniating tumor. Neurol Sci 22: 325–329 3. Chan YL, Leung SF, King AD, Choi PH, Metreweli C (1999) Late radiation injury to the temporal lobes: morphologic evaluation at MR imaging. Radiology 213: 800–807 4. Cheng KM, Chan CM, Fu YT, Ho LC, Tsang YW, Lee MK, Cheung YL, Law CK (2000) Brain abscess formation in radiation necrosis of the temporal lobe following radiation therapy for nasopharyngeal carcinoma. Acta Neurochir (Wien) 142: 435–440 5. Cheung M, Chan AS, Law SC, Chan JH, Tse VK (2000) Cognitive function of patients with nasopharyngeal carcinoma with and without temporal lobe radionecrosis. Arch Neurol 57: 1347–1352 6. Chong VF, Fan YF, Mukherji SK (2000) Radiation-induced temporal lobe changes: CT and MR imaging characteristics. AJR Am J Roentgenol 175: 431–436 7. Chong VF, Rumpel H, Fan YF, Mukherji SK (2001) Temporal lobe changes following radiation therapy: imaging and proton MR spectroscopic findings. Eur Radiol 11: 317–324 8. Gonul E, Izci Y, Safali M, Baysefer A, Seber N (2001) Multiple sclerosis with mass effect: case report. Minim Invasive Neurosurg 44: 117–120 9. Hartmann M, Heiland S, Sartor K (2000) Functional MRI procedures in the diagnosis of brain tumors: perfusion- and diffusionweighted imaging. Fortschr R€ontgenstr 174: 955–964 10. Jellema K, van der Meulen MF, Witkamp TD, Taphoorn MJ (2001) Brain tumor or stroke? Ned Tijdschr Geneeskd 5: 145 11. Kumar AJ, Leeds NE, Fuller GN, Van Tassel P, Maor MH, Sawaya RE, Levin VA (2000) Malignant gliomas: MR imaging spectrum of radiation therapy- and chemotherapy-induced necrosis of the brain after treatment. Radiology 217: 377–384 12. Lee AW, Ng SH, Ho JH, Tse VK, Poon YF, Tse CC, Au GK, O SK, Lau WH, Foo WW (1988) Clinical diagnosis of late temporal lobe necrosis following radiation therapy for nasopharyngeal carcinoma. Cancer 15: 1535–1542 13. Lee AW, Ng SH, Tse VK, Chiu HM, Thaw M (1993) Bilateral temporal lobectomy for necrosis induced by radiotherapy for nasopharyngeal carcinoma. Acta Oncol 32: 343–344 14. Lee AW, Foo W, Chappell R, Fowler JF, Sze WM, Poon YF, Law SC, Ng SH, O SK, Tung SY, Lau WH, Ho JH (1998) Effect of time, dose, and fractionation on temporal lobe necrosis following radiotherapy for nasopharyngeal carcinoma. Int J Radiat Oncol Biol Phys 40: 35–42 15. McKnight TR, von dem Bussche MH, Vigneron DB, Lu Y, Berger MS, McDermott MW, Dillon WP, Graves EE, Pirzkall A, Nelson SJ (2002) Histopathological validation of a three-dimensional magnetic resonance spectroscopy index as a predictor of tumor presence. J Neurosurg 97: 794–802 16. Morgenstern LB, Frankowski RF (1999) Brain tumor masquerading as stroke. J Neurooncol Aug 44: 47–52 17. Okamoto K, Ito J, Furusawa T, Sakai K, Watanabe M, Tokiguchi S (1998) Small cortical infarcts mimicking metastatic tumors. Clin Imaging 22: 333–338 18. Schaefer PW, Gonzalez RG, Hunter G, Wang B, Koroshetz WJ, Schwamm LH (2001) J Diagnostic value of apparent diffusion coefficient hyperintensity in selected patients with acute neurologic deficits. Neuroimaging 11: 369–380 19. Schwartz RB, Mulkern RV, Gudbjartsson H, Jolesz F (1998) Diffusion-weighted MR imaging in hypertensive encephalopathy: clues to pathogenesis. AJNR Am J Neuroradiol 19: 859–862
26 20. Sugita Y, Terasaki M, Shigemori M, Sakata K, Morimatsu M (2001) Acute focal demyelinating disease simulating brain tumors: histopathologic guidelines for an accurate diagnosis. Neuropathology 21: 25–31 21. Sundgren PC, Edvardsson B, Holtas S (2002) Serial investigation of perfusion disturbances and vasogenic oedema in hypertensive encephalopathy by diffusion and perfusion weighted imaging. Neuroradiology 44: 299–304 22. Tsui EY, Chan JH, Leung TW, Yuen MK, Cheung YK, Luk SH, Tung SY (2000) Radionecrosis of the temporal lobe: dynamic susceptibility contrast MRI. Neuroradiology 42: 149–152 23. Tsui EY, Chan JH, Ramsey RG, Leung TW, Cheung YK, Luk SH, Lai KF, Wong KP, Fong D, Yuen MK (2001) Late temporal lobe necrosis in patients with nasopharyngeal carcinoma: evaluation with combined multi-section diffusion weighted and perfusion weighted MR imaging. Eur J Radiol 39: 133–138 24. Yarnykh VL (2002) Pulsed Z-spectroscopic imaging of crossrelaxation parameters in tissues for human MRI: theory and clinical applications. Magn Reson Med 47: 929–939
Comments The authors have analysed the occurrence of misdiagnosis in a series of 174 operative procedures ‘‘performed for glioma surgery’’ during a two years period. Thus, the analysis seems to exclude craniotomies performed for other primary and secondary brain tumours and nontumour lesions (lymphoma, metastases, etc) which may also raise similar diagnostic difficulties. The paper, which is well written, is just a collection of nicely illustrated cases and does not define a new strategy for improving preoperative differential diagnosis of intracranial lesions. Both the analysis of the cases reported, and the revision of the literature relate to diagnostic difficulties posed by specific lesions such as brain infarction, multiple sclerosis and radiation necrosis, which are the most common lesions for misdiagnosis in daily practice. It is well known that despite increasing sensitivity and specificity of magnetic resonance imaging there are still cases in whom preoperative diagnosis is missed. After calling our attention to misdiagnosis in ‘‘glioma cases’’, the authors wonder to what extent increased costs linked to the used of a more sophisticated pre-surgical evaluation, namely new: MRI techniques, could be offset by avoidance of non-indicated craniotomies. This consideration is worth, but the question is whether or not these advanced techniques may solve at the present moment the problems of misdiagnosis they have analysed in this work. They recognize that further
G. Wurm et al.: Distinct supratentorial lesions mimicking cerebral gliomas research is still necessary in the field of modem imaging techniques (diffusion, perfusion and spectroscopy) to evaluate its value for the diagnosis of specific brain lesions. Thus, they could use currently available advanced MR techniques considered in the Discussion section in a prospective manner (as they did in the patient illustrated in Figure n 5) and then presenting their results. When considering the cases presented, I think that diagnoses of infarction and radiation necrosis should have been more strongly considered in cases n 1 and 4, respectively; consequently a ‘‘wait and see’’ policy in the first case, and a steroid trial in the second could have been adopted. Though in case n 1 they considered the alternative diagnosis of DNET (the lesion involved the cortex) the possibility of infarction was much more likely in my opinion. Perhaps, stereotactic biopsy or PET scanning could have avoided craniotomy in some of the cases presented, as occurred in case n 5 in whom biopsy showed multiple sclerosis thus avoiding improper conventional resection. R. Diez Lobato The authors analysed the occurrence of misdiagnosis in a series of 197 operative procedures ‘‘performed for glioma surgery’’ during a two years period. Both the cases reported and the revision of the literature relate to diagnostic difficulties posed by specific lesions such as brain infarction, multiple sclerosis and radiation necrosis, which are the most common places for misdiagnosis in daily practice. It is clear that despite increasing sensitivity and specificity of magnetic resonance imaging, there are still cases in which preoperative diagnosis is missed. After calling our attention to the problem of misdiagnosis, the authors wonder to what extent increased costs linked to the use of more sophisticated pre-surgical evaluation such as new MRI techniques, could be offset by avoidance of non indicated craniotomies. This consideration is worth, but the question is whether or not these advanced techniques may at the present moment solve the problems of misdiagnosis reported in this work. They recognize that further research is still necessary in the field of MRI imaging (diffusion, perfusion and MR-spectroscopy) to evaluate its diagnostic reliability. When considering the cases reported, I think that diagnoses of infarction and radiation necrosis should have been more strongly considered in cases no. 1 and 4, respectively. Ramiro D. Lobato
Correspondence: Dr. Gabriele Wurm, Neurosurgical Department, Landes-Nervenklinik Wagner Jauregg, Wagner Jauregg Weg 15, A-4021 Linz, Austria. e-mail:
[email protected]