British Journal of Neurosurgery
ISSN: 0268-8697 (Print) 1360-046X (Online) Journal homepage: http://www.tandfonline.com/loi/ibjn20
Is peroperative smear cytology necessary for CTguided stereotaxic biopsy? K. S. O'neill, P. V. Dyer, B. A. Bell, P. R. Wilkins, D. Uttley & H. T. Marsh To cite this article: K. S. O'neill, P. V. Dyer, B. A. Bell, P. R. Wilkins, D. Uttley & H. T. Marsh (1992) Is peroperative smear cytology necessary for CT-guided stereotaxic biopsy?, British Journal of Neurosurgery, 6:5, 421-427, DOI: 10.3109/02688699208995030 To link to this article: http://dx.doi.org/10.3109/02688699208995030
Published online: 06 Jul 2009.
Submit your article to this journal
Article views: 7
View related articles
Citing articles: 2 View citing articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=ibjn20 Download by: [137.189.171.235]
Date: 17 March 2016, At: 02:17
British Journal of Neurosurgery (1992) 6 , 421-427
ORIGINAL ARTICLE
Is peroperative smear cytology necessary for CT-guided stereotaxic biopsy? K. S. O'NEILL, P. V. DYER, B. A. BELL, P. R. WILKINS, D. UTTLEY & H. T. MARSH
Downloaded by [] at 02:17 17 March 2016
Atkinson Morley 's Hospital, Copse Hill, Wimbledon, London S W20 ONE, UK
Abstract Neuropathology is a relatively scarce resource with uneven geographical distribution, and some surgeons rely on macroscopic appearances of tissue and await later paraffin section histology for the final diagnosis. The effect of the introduction of a peroperative cytology service has been assessed in a 5-year series of CT-directed stereotaxic biopsies of intracranial mass lesions using a low cost adapted pre-CT stereotaxic frame. The technique has been mastered by 28 different surgeons with varying degrees of neurosurgical expertise and 259 procedures have been performed in 245 patients. Benign lesions were detected in 24 (9.8%) patients, confirming the importance of making a histological diagnosis prior to initiating treatment. Permanent morbidity was 6.5% and mortality 3.3% and morbidity was balanced by an improvement in 19.6% of the patients following the procedure. For the first 142 patients no peroperative histological diagnosis was available and a diagnostic rate of 86.6% was achieved. For the last 103 patients the availability of peroperative smears improved the diagnostic rate to 94.2% and reduced the number of second biopsies needed for diagnosis from seven to zero. Our series demonstrates the need for peroperative cytology in C T stereotaxic biopsies, and confirms that a complete neuropathology service is a prerequisite for neurosurgical patient care.
Key words: Brain tumour biopsy, computed tomography, smear diagnosis, stereotaxic surgery.
Introduction Neuropathology has an uneven geographical distribution within the UK, and worldwide is a relatively scarce resource. Whilst some neurosurgical units have excellent neuropathological backup, many units do not have a neuropathological opinion constantly available. Some surgeons have to rely on the initial macroscopic appearances of tissue and await later paraffin section histology for the final diagnosis. The accuracy of neuroradiology in intrinsic brain lesions has increased with advances in computed tomography (CT) and magnetic resonance imaging (MRI), but there remains a proportion of lesions which may be diagnosed wrongly and treated inappropriately. The sep-
aration of extrinsic from intrinsic tumours can still be difficult, even on MRI, and histological verification still remains the only means of obtaining a definitive diagnosis.'T2 Peroperative cytology gives the surgeon immediate guidance on the likely nature of the lesion, and determines whether further biopsies are needed to establish the diagnosis. A positive smear may also reveal the unexpected, and enable more extensive surgery to proceed under the same anaesthetic if appropriate. We present the results of C T stereotaxic biopsy using a pre-CT Leksell stereotaxic frame adapted to allow accurate CT-directed stereotaxic biopsy in combination with an IGE 9800 CT canner.^ At the commencement of our
42 1
L22
Y. S. O'Neill et al.
ieries peroperative cytology was not available o us, and its introduction has allowed us to mess its contribution to the diagnostic success )f C T stereotaxic biopsy.
latient and continued for an average o f 4 days lostoperatively .
Downloaded by [] at 02:17 17 March 2016
Materials and methods
Between October 1986 and July 1991, using :he system described previously,3 a total of 245 iatients (133 men and 112 women), with a mean age of 56 years (range 7-85 years), had 159 procedures performed by 28 different surgeons (Fig. 1). The first 142 patients did aot have peroperative cytology, and multiple iiopsy specimens were taken and sent in formalin to the laboratory for paraffin section. [n the subsequent 103 patients who had peroperative smears at each CT-determined Eoordinate, two biopsy specimens were taken. One was sent immediately for a smear, by the pathologist in a laboratory adjacent to the theatre, the other was sent in formalin for a paraffin section. If the pathologist was confident of the result of the first smear, further biopsies were not taken but if the initial smear was not diagnostic, further biopsies were taken at different coordinates until the smear was diagnostic or the surgeon deemed that further biopsies were inadvisable. Dexamethasone (4 mg 6 hourly) was given preoperatively in every
.,
1. Seniority of the surgeon performing the biopsy xocedure. consultant; a, senior registrar; m, regisrar; 0, senior house officer.
;IG.
Kesults
Biopsies were taken from eight intracranial regions, including 102 deeply placed lesions, and 17 posterior fossa lesions (Fig. 2). There were 86 right-sided lesions, 98 left-sided and 61 other sites which included midline and bilateral lesions. The majority of patients had intrinsic tumours and 146 astrocytomas were diagnosed at first biopsy. In 48 patients other malignant tumours were discovered including lymphoma in 13 patients (Fig. 3). Benign lesions were diagnosed in 24 patients. These included three meningiomas, one
m Biopsy site FIG. 2. Anatomical location of the biopsy site within the brain.
Peroperative smear cytology
423
:i 60
Downloaded by [] at 02:17 17 March 2016
Malignant histology
FIG.3. Distribution of histology of the malignant tumours.
craniopharyngioma, and one colloid cyst. Brain abscesses were diagnosed in 12 patients and drained appropriately, and stereotaxic biopsy was not only a useful diagnostic tool in these patients but was also therape~tic.~ All but two of our patients with abscesses improved considerably after stereotaxic drainage. Other non-neoplastic diagnoses obtained included toxoplasmosis in a patient with AIDS, tuberculous granuloma, and histiocytosis (Fig. 4). Fourteen patients had a second biopsy. This was performed to establish a positive diagnosis where the first biopsy failed to do so in six patients, to reconfirm a positive diagnosis where there was clinical uncertainty in five patients, and where the first biopsy had secured a diagnosis and drainage had proved therapeutic in three patients. The first biopsy failed to establish the diagnosis in 27 patients; in 16 of these the biopsy showed normal brain tissue, in eight necrotic material, and in three reactive gliosis. Whether these biopsies were diagnostic failures was determined by establishing a final diagnosis by either a second biopsy or by assessing follow-up radiological studies and clinical outcome. Of the 16 patients with normal biopsies, five were rebiopsied to reveal three astrocytomas, one teratoma, and one patient with organizing
fibrin whose final diagnosis was sarcoidosis. The remaining 11 patients with normal biopsies had a final diagnosis made on subsequent radiological appearances and clinical outcome. The final diagnosis was a tumour in nine patients, viral encephalitis in one, and multiple sclerosis in one. Of the eight patients with a biopsy that was necrotic, one was rebiopsied to show an oligodendroglioma, six more were diagnosed as tumours, and one as an abscess. The three patients whose biopsy showed reactive change only were finally diagnosed on radiological and clinical grounds as having multiple sclerosis, Wilson's disease, and a tumour, respectively. In multiple sclerosis and Wilson's disease reactive gliosis is all that may be found on histology, and these biopsies cannot therefore be regarded as negative (Fig. 5). The overall diagnostic rate of the procedure in the 245 patients was 89.8%, but the introduction of peroperative smear cytology in April 1989 changed the diagnostic rate of the initial biopsy. Prior to April 1989,142 patients had biopsies, and 19 were negative, yielding a diagnostic rate of 86.6%. After April 1989,103 patients had biopsies and six were negative, giving a diagnostic rate of 94.2%. The number of second procedures after the introduction of peroperative smears was only two, compared
424
K. S. O'Neill et al.
Downloaded by [] at 02:17 17 March 2016
Benign histology
FIG. 4. Distribution of histology of the benign lesions.
with 12 before, and these were both therapeutic-to redrain an abscess in one patient and a cystic tumour in the other (Fig. 6). An overall mortality rate of 3.3% was adhieved, and of the eight patients who died, two had grade IV astrocytomas, three had grade I11 astrocytomas, two had metastatic tumours and one had necrotic debris most probably from a malignant tumour. An immediate morbidity of 11.0% arises from 27 patients who were immediately worse after the procedure. Of these patients, histology revealed tumour in 23, gliosis in one, haemorrhage in one, haemangioblastoma in one and necrotic tissue in another. Some of these patients subsequently improved, the deterioration often lasting a matter of hours. The number of patients with a permanent deficit was only 16, therefore reducing lasting morbidity of the procedures to 6.5%. In those patients who died, 25% of the procedures were performed by senior registrars and 75% by registrars. Those patients who were worse immediately following a procedure had been operated on by senior registrars in 33.3% of cases, by registrars in 61.9%, and by senior house officers in 4.8%. These figures reflect the fact that most procedures were done by registrars, and the low morbidity following
procedures by senior house officers indicates the safety of the method. An improvement was seen in the condition of 34 patients immediately after the procedure. Another 14 patients were initially unchanged after the biopsy but subsequently improved. This represents an overall improvement rate of 19.6%, which considerably exceeds the morbidity rate (Fig. 7). Of those patients who improved, nine had drainage of an abscess stereotaxically, one had a tuberculous granuloma, one a meningioma, and the remaining 37 had malignant tumours.
Discussion It has been suggested that the majority of intracranial biopsies may be performed safely, accurately and economically with a CT-guided freehand needle.5 The biopsy is performed with the patient positioned in the C T scanner and the needle trajectory planned as closely perpendicular to the entry site in the skull as possible. The biopsy needle is passed through the brain until tumour is encountered; however, the precise site from which tissue has been removed freehand is unknown, a considerable disadvantage if the sample histology is negative. Freehand burrhole biopsy for deep lesions and those in the posterior fossa has
Peroperative smear cytnlngy Initial Histology
Diagnostic Method
425
Final Diagnosis
Rebiopsy (5j
ormal brain (15)
\
Tumour (8) Clinical outcome (10) Multiple sclerosis (1)
Rebiopsy (1)
Downloaded by [] at 02:17 17 March 2016
peroperative cytology (21)
Oligodendrogiioma (1)
ecrotic tissue (4 Clinical outcome (3)
I
\
Tumour (3)
ISSN: 0268-8697 (Print) 1360-046X (Online) Journal homepage: http://www.tandfonline.com/loi/ibjn20
Is peroperative smear cytology necessary for CTguided stereotaxic biopsy? K. S. O'neill, P. V. Dyer, B. A. Bell, P. R. Wilkins, D. Uttley & H. T. Marsh To cite this article: K. S. O'neill, P. V. Dyer, B. A. Bell, P. R. Wilkins, D. Uttley & H. T. Marsh (1992) Is peroperative smear cytology necessary for CT-guided stereotaxic biopsy?, British Journal of Neurosurgery, 6:5, 421-427, DOI: 10.3109/02688699208995030 To link to this article: http://dx.doi.org/10.3109/02688699208995030
Published online: 06 Jul 2009.
Submit your article to this journal
Article views: 7
View related articles
Citing articles: 2 View citing articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=ibjn20 Download by: [137.189.171.235]
Date: 17 March 2016, At: 02:17
British Journal of Neurosurgery (1992) 6 , 421-427
ORIGINAL ARTICLE
Is peroperative smear cytology necessary for CT-guided stereotaxic biopsy? K. S. O'NEILL, P. V. DYER, B. A. BELL, P. R. WILKINS, D. UTTLEY & H. T. MARSH
Downloaded by [] at 02:17 17 March 2016
Atkinson Morley 's Hospital, Copse Hill, Wimbledon, London S W20 ONE, UK
Abstract Neuropathology is a relatively scarce resource with uneven geographical distribution, and some surgeons rely on macroscopic appearances of tissue and await later paraffin section histology for the final diagnosis. The effect of the introduction of a peroperative cytology service has been assessed in a 5-year series of CT-directed stereotaxic biopsies of intracranial mass lesions using a low cost adapted pre-CT stereotaxic frame. The technique has been mastered by 28 different surgeons with varying degrees of neurosurgical expertise and 259 procedures have been performed in 245 patients. Benign lesions were detected in 24 (9.8%) patients, confirming the importance of making a histological diagnosis prior to initiating treatment. Permanent morbidity was 6.5% and mortality 3.3% and morbidity was balanced by an improvement in 19.6% of the patients following the procedure. For the first 142 patients no peroperative histological diagnosis was available and a diagnostic rate of 86.6% was achieved. For the last 103 patients the availability of peroperative smears improved the diagnostic rate to 94.2% and reduced the number of second biopsies needed for diagnosis from seven to zero. Our series demonstrates the need for peroperative cytology in C T stereotaxic biopsies, and confirms that a complete neuropathology service is a prerequisite for neurosurgical patient care.
Key words: Brain tumour biopsy, computed tomography, smear diagnosis, stereotaxic surgery.
Introduction Neuropathology has an uneven geographical distribution within the UK, and worldwide is a relatively scarce resource. Whilst some neurosurgical units have excellent neuropathological backup, many units do not have a neuropathological opinion constantly available. Some surgeons have to rely on the initial macroscopic appearances of tissue and await later paraffin section histology for the final diagnosis. The accuracy of neuroradiology in intrinsic brain lesions has increased with advances in computed tomography (CT) and magnetic resonance imaging (MRI), but there remains a proportion of lesions which may be diagnosed wrongly and treated inappropriately. The sep-
aration of extrinsic from intrinsic tumours can still be difficult, even on MRI, and histological verification still remains the only means of obtaining a definitive diagnosis.'T2 Peroperative cytology gives the surgeon immediate guidance on the likely nature of the lesion, and determines whether further biopsies are needed to establish the diagnosis. A positive smear may also reveal the unexpected, and enable more extensive surgery to proceed under the same anaesthetic if appropriate. We present the results of C T stereotaxic biopsy using a pre-CT Leksell stereotaxic frame adapted to allow accurate CT-directed stereotaxic biopsy in combination with an IGE 9800 CT canner.^ At the commencement of our
42 1
L22
Y. S. O'Neill et al.
ieries peroperative cytology was not available o us, and its introduction has allowed us to mess its contribution to the diagnostic success )f C T stereotaxic biopsy.
latient and continued for an average o f 4 days lostoperatively .
Downloaded by [] at 02:17 17 March 2016
Materials and methods
Between October 1986 and July 1991, using :he system described previously,3 a total of 245 iatients (133 men and 112 women), with a mean age of 56 years (range 7-85 years), had 159 procedures performed by 28 different surgeons (Fig. 1). The first 142 patients did aot have peroperative cytology, and multiple iiopsy specimens were taken and sent in formalin to the laboratory for paraffin section. [n the subsequent 103 patients who had peroperative smears at each CT-determined Eoordinate, two biopsy specimens were taken. One was sent immediately for a smear, by the pathologist in a laboratory adjacent to the theatre, the other was sent in formalin for a paraffin section. If the pathologist was confident of the result of the first smear, further biopsies were not taken but if the initial smear was not diagnostic, further biopsies were taken at different coordinates until the smear was diagnostic or the surgeon deemed that further biopsies were inadvisable. Dexamethasone (4 mg 6 hourly) was given preoperatively in every
.,
1. Seniority of the surgeon performing the biopsy xocedure. consultant; a, senior registrar; m, regisrar; 0, senior house officer.
;IG.
Kesults
Biopsies were taken from eight intracranial regions, including 102 deeply placed lesions, and 17 posterior fossa lesions (Fig. 2). There were 86 right-sided lesions, 98 left-sided and 61 other sites which included midline and bilateral lesions. The majority of patients had intrinsic tumours and 146 astrocytomas were diagnosed at first biopsy. In 48 patients other malignant tumours were discovered including lymphoma in 13 patients (Fig. 3). Benign lesions were diagnosed in 24 patients. These included three meningiomas, one
m Biopsy site FIG. 2. Anatomical location of the biopsy site within the brain.
Peroperative smear cytology
423
:i 60
Downloaded by [] at 02:17 17 March 2016
Malignant histology
FIG.3. Distribution of histology of the malignant tumours.
craniopharyngioma, and one colloid cyst. Brain abscesses were diagnosed in 12 patients and drained appropriately, and stereotaxic biopsy was not only a useful diagnostic tool in these patients but was also therape~tic.~ All but two of our patients with abscesses improved considerably after stereotaxic drainage. Other non-neoplastic diagnoses obtained included toxoplasmosis in a patient with AIDS, tuberculous granuloma, and histiocytosis (Fig. 4). Fourteen patients had a second biopsy. This was performed to establish a positive diagnosis where the first biopsy failed to do so in six patients, to reconfirm a positive diagnosis where there was clinical uncertainty in five patients, and where the first biopsy had secured a diagnosis and drainage had proved therapeutic in three patients. The first biopsy failed to establish the diagnosis in 27 patients; in 16 of these the biopsy showed normal brain tissue, in eight necrotic material, and in three reactive gliosis. Whether these biopsies were diagnostic failures was determined by establishing a final diagnosis by either a second biopsy or by assessing follow-up radiological studies and clinical outcome. Of the 16 patients with normal biopsies, five were rebiopsied to reveal three astrocytomas, one teratoma, and one patient with organizing
fibrin whose final diagnosis was sarcoidosis. The remaining 11 patients with normal biopsies had a final diagnosis made on subsequent radiological appearances and clinical outcome. The final diagnosis was a tumour in nine patients, viral encephalitis in one, and multiple sclerosis in one. Of the eight patients with a biopsy that was necrotic, one was rebiopsied to show an oligodendroglioma, six more were diagnosed as tumours, and one as an abscess. The three patients whose biopsy showed reactive change only were finally diagnosed on radiological and clinical grounds as having multiple sclerosis, Wilson's disease, and a tumour, respectively. In multiple sclerosis and Wilson's disease reactive gliosis is all that may be found on histology, and these biopsies cannot therefore be regarded as negative (Fig. 5). The overall diagnostic rate of the procedure in the 245 patients was 89.8%, but the introduction of peroperative smear cytology in April 1989 changed the diagnostic rate of the initial biopsy. Prior to April 1989,142 patients had biopsies, and 19 were negative, yielding a diagnostic rate of 86.6%. After April 1989,103 patients had biopsies and six were negative, giving a diagnostic rate of 94.2%. The number of second procedures after the introduction of peroperative smears was only two, compared
424
K. S. O'Neill et al.
Downloaded by [] at 02:17 17 March 2016
Benign histology
FIG. 4. Distribution of histology of the benign lesions.
with 12 before, and these were both therapeutic-to redrain an abscess in one patient and a cystic tumour in the other (Fig. 6). An overall mortality rate of 3.3% was adhieved, and of the eight patients who died, two had grade IV astrocytomas, three had grade I11 astrocytomas, two had metastatic tumours and one had necrotic debris most probably from a malignant tumour. An immediate morbidity of 11.0% arises from 27 patients who were immediately worse after the procedure. Of these patients, histology revealed tumour in 23, gliosis in one, haemorrhage in one, haemangioblastoma in one and necrotic tissue in another. Some of these patients subsequently improved, the deterioration often lasting a matter of hours. The number of patients with a permanent deficit was only 16, therefore reducing lasting morbidity of the procedures to 6.5%. In those patients who died, 25% of the procedures were performed by senior registrars and 75% by registrars. Those patients who were worse immediately following a procedure had been operated on by senior registrars in 33.3% of cases, by registrars in 61.9%, and by senior house officers in 4.8%. These figures reflect the fact that most procedures were done by registrars, and the low morbidity following
procedures by senior house officers indicates the safety of the method. An improvement was seen in the condition of 34 patients immediately after the procedure. Another 14 patients were initially unchanged after the biopsy but subsequently improved. This represents an overall improvement rate of 19.6%, which considerably exceeds the morbidity rate (Fig. 7). Of those patients who improved, nine had drainage of an abscess stereotaxically, one had a tuberculous granuloma, one a meningioma, and the remaining 37 had malignant tumours.
Discussion It has been suggested that the majority of intracranial biopsies may be performed safely, accurately and economically with a CT-guided freehand needle.5 The biopsy is performed with the patient positioned in the C T scanner and the needle trajectory planned as closely perpendicular to the entry site in the skull as possible. The biopsy needle is passed through the brain until tumour is encountered; however, the precise site from which tissue has been removed freehand is unknown, a considerable disadvantage if the sample histology is negative. Freehand burrhole biopsy for deep lesions and those in the posterior fossa has
Peroperative smear cytnlngy Initial Histology
Diagnostic Method
425
Final Diagnosis
Rebiopsy (5j
ormal brain (15)
\
Tumour (8) Clinical outcome (10) Multiple sclerosis (1)
Rebiopsy (1)
Downloaded by [] at 02:17 17 March 2016
peroperative cytology (21)
Oligodendrogiioma (1)
ecrotic tissue (4 Clinical outcome (3)
I
\
Tumour (3)
