CT-Assisted Stereotactic Brain Biopsy: Value of Intraoperative Frozen Section Diagnosis

J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.51.3.332 on 1 March 1988. Downloaded from

Journal of Neurology, Neurosurgery, and Psychiatry 1988;51:332-341

CT-assisted stereotactic brain biopsy: value of intraoperative frozen section diagnosis

HAROLD J COLBASSANI,* SHUNJI NISHIO,t KEVIN M SWEENEY,* ROY A E BAKAY,* YOSHIO TAKEIt From the Departments of Pathology (Neuropathology) t and Surgery (Neurosurgery),* Emory University School ofMedicine, Atlanta, Georgia, USA

SUMMARY In 100 recent CT-guided brain biopsies, the value of intraoperative histologic examination using frozen section technique was evaluated. In 87 of these cases, the biopsy was performed stereotactically. In the remaining 13 cases, a CT-guided free hand technique was used. Of the 100 biopsies performed, adequate tissue for histopathologic diagnosis was obtained in 97, and in three the biopsy was nondiagnostic. In 61 procedures the initial biopsy specimen was adequate for diagnosis. Two specimens were required in 25 and in the remaining cases it was necessary to obtain three to four biopsy specimens before a definitive diagnosis could be made. Ultimately, the histologic

diagnosis was made on frozen section examination in 93 of the cases. The lesions identified were Protected by copyright. neoplastic disease in 83 cases, vascular disease in seven, infectious disease in five, demyelinating disease in one, and radiation necrosis in one. Comparison between the frozen section diagnosis and the final diagnosis based on the permanent sections revealed that they matched in 89 cases (92%). Of the 83 cases of neoplasms the exact grade of malignancy was determined by frozen section examination in 71 (85%). Comparison between the size of tissue samples obtained and the ability to make a final diagnosis revealed that even if the specimen volume was less than 2 mm3, the biopsy was generally successful. The disadvantages of the small sample size obtained through needle biopsy are best overcome by careful targeting and assessment of sample quality by intraoperative frozen section examinations, which will give the definitive diagnosis in most of the cases without paraffin-embedded sections.

Despite the high resolution of computed tomography in highly functional areas both safe and reliable.8 -1 (CT), and magnetic resonance imaging (MRI), the While stereotactic biopsy increases the accuracy of http://jnnp.bmj.com/ specificity of these modalities remains limited.`6 obtaining appropriate tissue for precise diagnosis, Therefore, therapeutic decisions must be based upon specimens obtained by the initial needle passage are an unequivocal histological diagnosis. sometimes insufficient for a definitive histologic inter- In 1947 Spiegel et al first introduced the use of pretation. Microscopical examination of such stereotactic techniques in obtaining biopsy material specimens provides the ability to determine the neces- from human subcortical structures.7 Today, these sity of additional needle passages. Thus, rapid, intra- basic principles remain the same; however, the operative histological examination of each biopsy combination of stereotactic techniques and computed sample by a well-trained neuropathologist will pro- on September 29, 2021 by guest. tomography has made biopsy of even minute lesions vide complete and reliable information on a minimum amount of tissue and avoid the risk of further biopsy. We review our experience on 100 brain biopsies, in Address for reprint requests: Yoshio Takei, MD, Division of which a CT-guided needle biopsy and intraoperative Neuropathology, Emory University Hospital, 1364 Clifton Road, frozen section examination was performed for N.E. Atlanta, Georgia 30322, USA. diagnosis of a variety of pathological entities. The Received 16 March 1987 and in revised form 3 June 1987. value of intraoperative frozen section diagnosis in Accepted 14 July 1987 CT-guided brain biopsy is evaluated. 332 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.51.3.332 on 1 March 1988. Downloaded from

CT-assisted stereotactic brain biopsy: value of intraoperative frozen section diagnosis 333 Table 1 Location ofbiopsied lesions (N = 97) Table 2 Final diagnosis ofbiopsied lesions (N = 97)

Supratentorial 86 Neoplastic lesion 83 Cerebral hemisphere Astrocytoma Left 39 Fibrillary 18 Right 1 5 Protoplasmic 9 Corpus callosum 5 Gemistocytic 2 Basal ganglia 4 Malignant 23 Thalamus, hypothalamus 16 Glioblastoma multiforme 7 Pineal region I Other gliomas* 4 Ventricle Neuroblastoma 2 Lateral 2 Malignant lymphoma 2 Third 4 Other primary brain tumourt 5 Metastatic tumour 11 Infratentorial 11 Vascular lesion 7 Midbrain 3 Haematoma 2 Pons/medulla 6 Infarction 5 Cerebellum 2 Infectious lesion 5 Total 97 Others$ 2 Total 97 *Ependymoma (I); choroid plexus papilloma(1); oligodendroglioma (1); subependymoma (1). Patients and methods tHamartoma (1); germinoma (1); craniopharyngioma(I); neuroepithelial cyst (1); pineoblastoma (1). During the period April 1983 to June 1986, 100 CT-guided $Multiple sclerosis (1); radiation necrosis (1). needle biopsies of brain lesions were performed on 98 patients at Emory University Hospital. In 87 cases, biopsies The final pathological diagnosis was the result of rapid were performed with CT guidance (General Electric 8800 intraoperative frozen section examination, histological ex- scanner), using the multiplanar stereotactic head frame amination of paraffin-embedded materials, and sometimes

(Howland Industries, Stanton, California) and software electron microscopic or immunocytochemical examinations. Protected by copyright. to produce simultaneous axial, coronal, and sagittal This diagnosis was compared with the clinical impression reconstructions, allowing a three-dimensional assessment of and frozen section diagnosis in each case. Special attention the target tissue.9 In the remaining 13 cases the biopsy was was paid to any discrepancies and these were reviewed in performed by a free hand technique. The biopsy procedure conjunction with an assessment of sample volume and type was generally carried out under local anaesthesia, except in of lesions. children younger than 12 years. The patients included 53 males and 44 females, aged from I to 89 years (mean Results 37 5). The anatomical origin of the specimens is given in table 1. After combined CT-guided biopsy and intraoperative Specimens were obtained with an 18-gauge guillotine frozen section examinations, an adequate sample for needle, usually from the periphery or contrast enhancing histological confirmation of lesion type was obtained edge (if present), and the centre of the lesion shown on CT scan. Specimens were placed on a flat piece of rubber glove in 97 of the 100 biopsies. In the three nondiagnostic and then wrapped in a towel to keep them moist. The specimens were examined immediately by the pathologists and Table 3 Correlation ofclinical impression withfinal the approximate volume of the sample was evaluated in histopathological diagnosis (N = 97) many cases. Frequently, a small sample was taken for electron microscopy, and then the remainder of the specimen Agreement withfinal http://jnnp.bmj.com/ was frozen and sectioned in a cryostat (American Optical histopathological diagnosis? Cryostat, Buffalo, NY), which was kept at -20°C. A small mount comprising ofembedding medium for frozen sections Clinical diagnosis Yes No was prepared on a chuck inside the cryostat and then frozen Primary brain tumour 60 9* with the aid of a commercially available freon spray. After Primary vs metastatic tumour 7 0 freezing, sections were cut at 3 or 4 um thickness. The slides Primary tumour vs radiation were then stained with haematoxylin and eosin, and tolu- necrosis 3 0 Tumour vs infectious lesion 7 0 idine blue (0 7% solution). The frozen specimens were sub- Tumour vs infarction 4 0 on September 29, 2021 by guest. sequently fixed in 10% formalin solution and were processed Infarction 2 0 through a routine paraffin embedding technique. The histo- Otherst 4 1$ pathological findings of the frozen section, and sections were the senior *Included is a case in which the clinical diagnosis of "brain tumour paraffin-embedded interpreted by ofglial origin" was identified histologically as "organising author (YT) in each case. Depending on the frozen section haematoma". This was subsequently confirmed by the patients' diagnosis and CT findings, one or more tissue samples were clinical course. usually then taken, until frozen section was estab- tHaematoma (1); abscess vs infarction (1); multiple sclerosis (2); diagnosis encephalitis (1). lished. In some cases it was decided to wait for tln this case a clinical diagnosis of "multiple sclerosis" was paraffin-embedded sections rather than subject the patient to identified histologically as "gemistocytic astrocytoma arising in a the additional risk of further biopsy. demyelinating focus". J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.51.3.332 on 1 March 1988. Downloaded from

334 Colbassani, Nishio, Sweeney, Bakay, Takei

Fig 1 (A) MRI scan ofa cystic lesion ofthe midbrain in a 12 year old boy. TE330 ms, TR 540 ms. Based upon the clinical impression and radiographicfindings, a working diagnosis ofa brainstem glioma was made. (B) CTscan obtained at the time ofstereotactic biopsy. The axial image as well as reconstructed coronal and sagittal images are seen. The simulated trajectory between entrance and target points is also depicted. (C) Biopsy specimen examined byfrozen section demonstrating a neuroepithelial cyst. This tissue was lost in the paraffin-embedded materials. Hematoxylin and Eosin, x 200. Protected by copyright.

biopsies, the procedure was terminated before an neuroepithelial cyst ofthe midbrain, and pineoblastoma. adequate tissue sample for confirmation of lesion type was obtained. However, these three biopsies 2 Correlation ofclinical impression tofinal were performed early in our series and it was thought histopathological diagnosis (table 3) additional biopsy would expose the patient to an A clinical impression of a primary brain tumour was unacceptable risk. Subsequently, a more aggressive confirmed histologically in 60 cases. In eight cases approach in obtaining tissue for a definitive diagnosis histological examination altered the clinical diagnosis has been adopted without increased morbidity. as to the tumour type. This included four metastatic Analyses mainly on the 97 biopsies, in which a tumours, which were diagnosed clinically as malig- definitive histological diagnosis was obtained, are nant astrocytoma in three cases and glioblastoma described below. multiforme in one case. Two neuroblastomas and one lymphoma were diagnosed clinically as malignant as- http://jnnp.bmj.com/ trocytomas; and one neuroepithelial cyst was initially diagnosed as a low grade astrocytoma (fig 1). In one I General aspects other case the diagnosis had to be changed to "or- The diagnoses based upon the histological findings ganising haematoma" which, clinically, was sus- are presented in table 2. The majority of the lesions in pected to be a brain tumour ofglial origin. Organising this series were tumours of the glioma group haematoma as a definitive diagnosis without evidence (N = 63), with the malignant astrocytoma being of neoplastic growth was confirmed by serial CT on September 29, 2021 by guest. the most frequent (N = 23). An ependymoma, scans and the subsequent clinical course of this pa- oligodendroglioma, subependymoma and choroid tient. A clinical impression of "primary vs secondary plexus papilloma were each encountered once. tumour" was found to be a primary brain tumour in Metastatic tumours were less frequent and were all seven cases. A clinical impression of "tumour vs identified in 11 cases. Two neuroblastomas, corro- radiation necrosis" was found to be tumour in two borated by ultrastructural examination, and two and radiation necrosis in one. Seven cases with a malignant lymphomas, corroborated by immuno- suspected diagnosis of "tumour vs infectious lesion" cytochemical studies were also seen. Rarely were found to be tumour in four and infection in the encountered were craniopharyngioma, germinoma, remaining three cases. J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.51.3.332 on 1 March 1988. Downloaded from

CT-as.sisted stereotactic brain biopsy: value of intraoperative frozen section diagnosis 335

I ...... ~~a......

-~~~~~~~~~~~~~~~~~~~~HI I Protected by copyright.

| ,;5:ki,0 ;;@ 0* F ej;0, I r S i t<00 iXAw

F~ ~ ~0

| ~3,,,S> ~ 0._< 'B 4t!0

|_eeqWtSX~~~~~~~~~~~~~~~~~~~~~~~~~~O http://jnnp.bmj.com/

k ;.X.iWai w AP on September 29, 2021 by guest. J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.51.3.332 on 1 March 1988. Downloaded from

336 Colbassani, Nishio, Sweeney, Bakay, Takei 3 Number oftissue samples Table 5 Sample volume and type oflesion (N = 49) There were 93 cases, in which the numbers of biopsy Mean sample volume samples were evaluable (table 4). Among 80 cases in Type oflesion (mm3) (range) which a stereotactic biopsy was performed, the initial specimen, which was usually obtained from the Benign astrocytoma (N = 19) 2-3 (05-12) ofthe lesion delineated on CT was Malignant astrocytoma/glioblastoma periphery sufficient multiforme (N = 23) 3-7 (05-20) for exact histological diagnosis in 51 cases (64%). A Metastatic tumour (N = 7) 3-5 (05-10) negative report on the initial sample necessitated an additional specimen in 21 cases. In six cases, three specimens were necessary, and two patients had four or more specimens taken before obtaining an appro- (85%) (fig 2). In seven cases, the frozen section diag- priate biopsy sample for histological diagnosis. nosis had to be changed in tumour type or grading Among the 13 cases, in which free hand biopsy with after examination of the paraffin-embedded sections. CT guidance was performed, the initial specimens This included four malignant astrocytomas, which were sufficient for histological diagnosis in six cases were diagnosed as "fibrillary or protoplasmic (46%). In two cases, two specimens were necessary, astrocytoma" on frozen section; one glioblastoma three patients had three specimens, and two patients diagnosed as a malignant astrocytoma (fig 3); one had four specimens taken before a histological diag- malignant astrocytoma, which was diagnosed as nosis could be made. "malignant neuroglial tumour" on frozen section; and one "malignant tumour" (no details concerning the tumour type could be determined on frozen 4 Volume ofsample section), which had to be changed to metastatic The volume of sample, from which a definitive carcinoma. histological diagnosis was obtained, was evaluable in A diagnostic error in frozen sections concerning the Protected by copyright. 71 biopsies. In 48 biopsies (67%) the volume of sam- presence of tumour tissue (false-negative diagnosis) ple was less than 2 mm3. The sample tended to be occurred in three cases. This included one malignant bigger in cases of metastatic tumours, malignant astrocytoma, in which the specimen was estimated as astrocytomas and glioblastomas than in those of low 5*5 mm3 in total sample volume and showed only gli- grade astrocytomas. However, there was no definite osis and oedema on frozen section; one glioblastoma correlation between the volume of samples and type multiforme, whose frozen section specimen revealed of lesion (table 5). The total sample volumes of the only necrotic tissue and was nondiagnostic. The total three cases in which the biopsy was nondiagnostic, sample volume of this latter biopsy was estimated ranged from 4 to 17mm3. as 6-5 mm3. Lastly was a case of a gemistocytic astrocytoma arising in a demyelinating focus, in which the frozen section showed only the evidence of 5 Accuracy ofintraoperative frozen section diagnosis demyelination (fig 4). Multiple sclerosis was the A correlation of the diagnosis obtained from frozen clinical impression of this patient. section and that obtained from paraffin-embedded In two biopsies, the intraoperative frozen section sections was carried out in 100 cases (table 6). Among diagnosis could not be confirmed in the permanent the 97 biopsies in which a definitive histologic diagno- sections since the lesions identified by frozen section http://jnnp.bmj.com/ sis was obtained, the diagnosis based on the frozen were lost in the paraffin-embedded materials (fig 1). In sections was confirmed by the examination of the these cases, the frozen section diagnosis had been postoperative paraffin-embedded sections in 89 relied upon for practical purposes, and this was biopsies as to tumour type or type of lesion. confirmed by the patient's clinical course. In the 83 cases of neoplasm, the exact grade of Among the biopsies of non-neoplastic lesions, the malignancy identified by frozen section examination frozen section diagnosis had to be changed from was confirmed by paraffin section examination in 71 astrocytic gliosis to multiple sclerosis in one case. Frozen section of this case showed astrocytic gliosis on September 29, 2021 by guest. Table 4 Number ofbiopsy specimens necessaryfor without evidence of demyelination, but the diagnosis paraffin-embedded section contained a demyelinating focus. Other non-neoplastic lesions were correctly 1 2 3 4 diagnosed by frozen section. In the three nondiagnostic biopsies, frozen and Stereotactic biopsy (N = 80) 51 21 6 2 paraffin-embedded sections contained only Free hand technique (N = 13) 6 2 3 2 astrocytic Total (N = 93) 57 23 9 4 proliferation, suggesting the border zone of a pathological process. J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.51.3.332 on 1 March 1988. Downloaded from

CT-assisted stereotactic brain biopsy: value ofintraoperativefrozen section diagnosis 337 i.. V _.. '. _ :^ _^.A br ..: _ I 4. :.- ;w,.I I' '4,

:K.~~~~.A .*

8~~~~~~~~~~~~~~~~.4 * ** p * > r. .r.wL. 4 4 anr * 4%.4t4 :s F .e 4 4 ... Aw sA, *'St^sx .0 .. ..*#s a w _, * ._ 0i .,>^Ff .:s 9 _w

S :..ak01 V <. '7.: VS... S/. A_;~ ,#.. VW ..o. *e .4P Protected by copyright. is I._ 1 A.s ef Id- 9 WONIA. _

., .41i.1.- -." V,:'l.-l.. ,W.,. A.-A .j -Amma ,.CN..C... 'q*' 4 .. .- 7%E .:Y. %..Jmlkc.jy.: -33-,S:-. I A." :,_, Q -::i- 1: ..a,. 9... 14. k p I IWW. J::b nu

'.7' .4I. .,.....: ... st.. aA http://jnnp.bmj.com/

. .'"qw. # :-* .: % it. I* r jg I F 3 + it t~~~~~~~~~~i7iI I& ' .3 + i::i .... on September 29, 2021 by guest. .: .: **4 ,; 3-.^''

::,

e., % p: .'s A v . 0 *eR".. #iF P ...P _ .""WI' kr- 1:.: :sf. IAI 0 -A; Fig 2 Biopsy specimen ofa malignant astrocytoma, in whichfrozen andparaffin-embedded sections were in agreement. Nwnerous pleomorphic and multinucleateforms can be seen. (a) Frozen section. (b) Permanent section. Hematoxylin and Eosin x 900. Iti.X..o J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.51.3.332 on 1 March 1988. Downloaded from

338 Colbassani, Nishio, Sweeney, Bakay, Takei Discussion stereotactic technique and precise intraoperative frozen section diagnosis has been most useful in The indications for CT assisted stereotactic biopsy eliminating the undesirable (that is, nondiagnostic) are well-established. Nevertheless, while this outcome of the biopsy procedure. procedure has been shown to be safe and particularly As a means of making a rapid intraoperative effective for deep-seated lesions, there are several diagnosis, smear preparations have long been used, factors which influence its reliability. The first and particularly among pathologists of the European probably the most important factor is the selection of school.'5 19-26 Eisenhardt and Cushing22 and the target for biopsy.8 10-18 Although biopsy from a subsequently Badt20 were among its early advocates; contrast-enhancing rim will most often yield a diag- however, its use became widespread only after the nosis, several investigators have noted that viable introduction of the wet film technique by Russell et tumour tissue may be obtained from areas of al.26 The principal advantages of the smear technique enhancement and areas of low density.'0-2 16 In are its technical simplicity, the ease with which several addition, a diagnosis ofmalignant astrocytoma based tiny fragments of tissue can be screened, and the upon a biopsy of the contrast-enhancing rim may be clarity of detail. Smear preparations, however, often confidently diagnosed as a glioblastoma if necrosis is destroy tissue architecture, and are variable in thick- found within the low density centre. In our series, the ness so that the degree ofcellularity may be difficult to initial tissue sample was usually obtained from the evaluate. McMenemey emphasised the difficulties and periphery or contrast-enhancing rim (if present) of dangers of misinterpretations particularly when the lesion delineated on CT. This was checked histo- attempting to grade astrocytomas; as only 66% of logically by frozen section preparation and was found diagnoses based on smear preparations were to contain the lesion in 61% of cases. Thus, in 39% it confirmed by paraffin-embedded sections in his was necessary to take two or more subsequent series.23 More recent reports have shown a much samples before the definitive histopathological diag- higher degree of accuracy which in part is related toProtected by copyright. nosis was made. The second factor which influences improved stereotactic techniques. Marshall et a!24 the reliability of CT-guided biopsy is the size of the reported a correct diagnosis in 94% of cases and specimen sent for histological examination. This Ostertag et al'5 has had success with the smear tech- small sample size has been considered to be not only nique in 95% of his cases. Unfortunately, the number the essential limitation in making a diagnosis but also and volume of biopsy samples required for diagnosis the most common cause of misdiagnosis.19 As dem- were not specified in these articles. Interestingly, in onstrated in our series, however, the accuracy ofdiag- the report by Ostertag et al, a diagnosis based upon nosis does not always increase proportionately with paraffin-embedded sections was not possible in 31 % the size of the specimen. A sample volume of 2mm3 of the cases.'5 proved to be sufficient for diagnostic evaluation in Despite the reported accuracy of the smear 67% of cases. Clearly, these factors are not mutually technique, a definitive diagnosis can be obtained exclusive. The pathologist is dependent upon the neu- intraoperatively utilising the frozen section rosurgeon to obtain an optimal specimen for histo- technique." 13 1427 In the series by Bullard et al, logical study and the neurosurgeon must rely on the diagnosis was established on frozen section exam- expertise of the pathologist to inform him of a diag- ination of the first biopsy specimen in 65% ofcases." nostic biopsy. Thus, the combination of CT-assisted Additional specimens were necessary in the remaining http://jnnp.bmj.com/ cases before a diagnosis could be made. However, in Table 6 Correlation offrozen section diagnosis with approximately 50% of cases in which the first biopsy paraffin embedded sections (N = 97) sample was non-diagnostic on frozen section, the paraffin-embedded section examinations proved to be Positive Negative diagnostic. Ultimately, a diagnostic biopsy was Type oflesion correlation* correlation obtained in 49 of 50 cases. In our series, among 83 neoplasms, histological diagnoses based upon frozen Neoplastic (N = 83) on September 29, 2021 by guest. Presence of tumour 78 5t sections matched with the final diagnosis in tumour Type of tumour 76 7 type in 78 cases (94%), and the exact grade of malig- Grade of malignancy 71 12 Non-neoplastic (N = 14) nancy in 71 (85%). It should be noted, however, that Type of lesion 13 1 in neurosurgical pathology, one must be cautious in malignancy a tumour based upon the *Frozen section diagnosis was confirmed by paraffin embedded grading the of sections. histological appearance of limited samples alone. tlncluded are two cases, in which tumour was present only in the Clinical data, such as the age ofthe patient, the tempo frozen section, and three cases in which tumour was not present in the frozen section but was identified in the paraffin embedded and duration of symptoms, the radiographic findings sections. and the precise anatomical location of the lesion are J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.51.3.332 on 1 March 1988. Downloaded from

CT-assisted stereotactic brain biopsy: value of intraoperativefrozen section diagnosis 339

fa) > .~~~~~~~.....W..

/ =~~~~~~~~~~~~~~~~~~n.w= s _ e k E .. 4~~~~~~~~~4

IV~~~~~ .i~~~~~~~~~~ ~ ~ ~ ~ ~~:: ~ ~ ~~~~~~

*AW:~~.,.~~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ t ft~~~tw Protected by copyright. p~ ~~~~

S~~~~ ~ ~~~~~~~~~~S.# "?. -ii i..V

*e.3 I u;drr ^~~~~~~~ ¼ '4~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~4

ow.~,.,~~,. ~ ~ ~ _.,~ ~ ~ ~~.. 4. 7 L . ..- ~ . ~ *--F;| .gW .i' '- ,¾ http://jnnp.bmj.com/

:. i

. ;,.

14~~ ~ ~ ~ ~ ~ ~ ~ ~ on September 29, 2021 by guest.

*s '; ii' iY

(aFoensctondpitng o a.ig t F . y :

demonstrating a zone ofnecrosis and surrounding pseudopallisading, making the diagnosis ofa glioblastoma multtiforme more appropriate Hematoxylin and Eosin x 60. J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.51.3.332 on 1 March 1988. Downloaded from

340

".: -s :-: ..: ...4.: .'--....2'....,,.1: - ".. -..: .I....,O'..,.., .Y. .:ij;-:-:,:_--,,.#5111-.kl.".I..J ,--..I"' ....".,,::::... .% -`- 1- ,-:' ::: 1.%.. .:.: ....:.: ... :1,"' i"', ",::`.. :24 11 ,:;:.:S. .. ." ... -I-:-im, :: ...;-1V--'s1-,;:.4 -`&,:: ....:j ..,.,.. v:]:k::: :.. A: .- .-..",-':...,"..N.:"'.::. .' .-A -.;:::..- ..?-,::: N-1: ...::-.:.__. ..M-'O , W-i.,N-4A. ..M.. .:t-:...:..,...... I:...,:4It.. :fi:...... M..Y'AL..:. ".. W-10s,... .i. .:.-...... 1 ...... 4 :, .., -V -:11" ..:...... :.x `... :: ...i .:f: -ta....:, .. .." ..,,",....,W..: .:.',zz!:::.M::-..;.-.J."I...`1 "' .%. ..:.:.;,....4w...:-..41` ...... -.I11 .!h.i:.:W...:.:.I.-".64..F, " :.*..,*`i` ... -..!--,!- :5 .AV.'V.:.:: :u.::-...:'.. ,.:::.-. ..:."."...,,]i:; -.'.:".i."".W k_,-.:..,.4,""', "', .":.fy'. .:.: ::...... "" ...._!:..i- i;.',...... -.-:.,-s .3.....::::::. ,,..'."" :, .1-.. (..,!L) ..-."A. ..:0:4.,:;,,:..4?: .::-,: , -- .,...... i:::.....h..),,-i..f...0 "-...... :...... ,-M :1 :. ..0... .4-i-.-,i.._. ....c.. ::.. ...-.:: i4.. W 1', T...w.. ...,.":;I :-: m- .:.X:,... .;zl....'..,...,.iiH--x-,j.; ,.11.1...N.....X10....i.1.. A"....:.6 ,jT-,..:.:.l-.1 .: i-:,- .!,"::.,.."'. ff' IF,:...::..:.:",..:.....:%p. xI"...:-..: ..- ., $,t...`...- .a.:.v.' ".,..: ..,.- ..:..1.::- :.. .h...!:.V.:,.:,"",I..:.- ...... T:-_.. 4..,r-?-"'.VI11: :W.:...:.2:X:-..: ..,.:s: ..xI,I.--F.:X ,.s%. :. .:...... ,---4:::.:.::. Y...... I.1.!.:":,...... :I...:X.:.I.1.::.. .-I..: ::.... k.."'..,.I ". .,-:: .,...: :.- ..,:...... NA.Y_ :: -:.: ..:..'I.:::a. ...:.':...::!,.-f'. -11:'. ..i ...... I. .., ::1 ..:,..:.t.,4:_,: ":!% 1'. ... --. .i` .4..-.:...:...:...... "'kOl'...1", .': ..,.,. :. ,'': M. .-,....:"...,:A&,.,.... 'K.X:::...... :.. .-.i ..-.--, .- ....4...-.....:.--:,:. .;.:-.,:.....: .:I...:i.4iI.:11..N... Aoff- :1,'.::.. MR ..I-'1:.:i:.V..::4....;:...I..-43:N.:...... -1..:...-, "...i .:.1-::.:,.: '- -: -. .:..4. :,....t:::.''11., N A...... 4r .:.. :- 9.:X. .:1 .s. M..- :1 -:.1! ...... : i, .-:4...... ,M :. :. ,, C' INI...4..erl- ..%..-::.:...I....i:..4...... :.%:: .::..:...::, f.,..4% ...A....:.rI.:I--.- W. .,; ... 44. ..,.....:.:il... -Y ....'M .1.....:-- ji:.:..:% .:..-...... -....: %.N.4...... i_.:...1i ... :- 1. 1. :,..,. ...-:,: A t.:4.::...... %...... %.. .n:,;.::.....: 'N.... ::.;.....':.%:`1.- .::. ..1,,.....::1 .:. T, .5 :.....V. .:-: ..1. ..p.-....&PM: -.j:.:'N .. :....-. .R::fl.-..:-1:.....:I.":".,---,1,-,:.::;I*I':-,"-Ix--.. .:":-.,.: .1h .:.:: .I "i:...WX...... :. tOW.-5 .: :,k,::"P.1"Ci:.W-."... -1i...,...:- ..4 .:,.::!...... t`.:i:... ,-... ,,",,-..,:.:.....4..:..,-..t:...,..IJ.t ;s-,: ..4:s.., .....:A, ". 'N. "', --X 4.L....I'..1,1- "14 .:'' .i .:N...... !:. :.I.::0TA1iW :::,.. ,,... '... VO.V..,0::!.X- ....:.-:..A. .4 -..% ill:.,:.- -.--'14-- ":-:..,.....":7, ..::::,..; 11 -gii-",.: fili*-I.,!...... :1....,..: I -..:4,...... 4.:..., ..- --F:..:,4 - -X.:.....0,.:1,RF"",j:'..,..-`:.:4 ..:-,:.... - .:!!: *...... !:n..:!. .'M....:c...... '..:. 'N I.:. .:. ,:- .N ;..: "" :;..,.":..: -I." .k...,. .: 4' .. :.-.I..ie.N:::...-,z:..:..:;..20 ., .:..:. .?:. ...:. Olk.-..N.-:....F,-:-",,,.4 4: .i .:,, N.M.:,A..;.: ..4:- A.:W....I:.I,k...V.... -4x`,...... X:1-..-.,i.--j..,..::..,. X .,'! 1%...!'J, ,4 :1. '. -s ...... -:...... 1.l-, _:.:....4!":: -Z ..:-, .M.: ....:.....-"...4...-..,-..:.;:,,..1-..A11-X:I.: W.:..,:..:.. -- XV....:, "N .M4.i.:.."....1.'M ....X.-.-.-:.... ::. ,:1. ... .4- .:... "., ...,::-,&.:,.if.i.-...::: .....::t.," :,.. ...x:,A.....;.I...,...:::.. ,:I, -.... ':.. -HA.-.:."..-:...:.....::H,.",.:z:'j.:.:: ..,.1,-.:N-8M. -..... ::u...M"MW .::.:..'4:-.--....-....::.--. -::. -, ..,! N. ,...... 0 :,i '.,. 4..,. :..%.M...... 0..'"..:...A-14...... M :". :. N.. X.: .: ...... -::.:.-.,Ij!I. '-- .: .., 1.I -..'': --xf N. .I.?x.:.II.-t"' .. .:31'.11.. -11- :. .. :. s...... :.. .1., 'Ar:.:..-c ...:....-A:t!.I...:--- :4 .i .-N. ..,-...... %..:.....:.:I.:..1, ., :-,.,,."lfl:::....X:..,"".::N..-4''".-F..",:4,..I... :. .4:V -4:. ,::. ::.:,...:-. .-..-...... i....I.:..'k :Z.'-"' .:. I... ..is .F -.:-4-::%::.::.-:":...X.:': ::u:.s, 'si .. ".-:,.. :.,.::.,X..,".,.--t- -.:. I.: :.:,:.,..:...-1...... `--.--.....:-ir ...:A.:: ...."-4-.ih:....:.....,. -,.1%X.N11..:--:. .%:....I..?.,II..: ...:1 ....I...:...1...:..,."A :;..:: ....-.1...:hs:-.:...... -... --: :.:.. .m .x::. -..':-...:,.1:-..:... .:...I. 1.f:::...:.....:...4:4-- .:.W.4 .s-.i:: ....-:..4.N:_h.:.", .,. I!, .: A._,j.! %:-..I ..NM,:..;.. !::. .-...:%i:;:!-...... -.1.,.h ;;:- ...".-...,-,-.:,iAN,...... ,,:.....::.:YIL..-.: ...... i:.'..lZO.-V"'s..., ,4'%---!..!;jI -1 .:.%_4:.:,:...--,,.s....-X..-;.P:: .::I ..:...... :.. ...N...-...",i:,.....""...:.....:.:.;..;.t -..-:...::. ,,N:...--.4..-.:::. .,.1W. 1"_ ..-I !:: .1: -, .....5'..`...,'.....-.' "'h,.. .., .I.:-..I,Y..,_4%.,..N, .:. .i :i. "..:e:: ,..4-.: ,,, .A.ow.....--:::." -.s1i!-..:,-_-1W, "-,..Z.:...:.... :%.N.;-IC --p--4.1..: .i,--flli.--, ...:-!:X,.-.....%:- :...-'--.. --..::.,:.."N' .. .-.:.::: -,-,",....&_i -0. ...,.W.k...I....-.-:...:....:. ..,,ITF-i...;.il.,5-i...:... t.- ...-P:.V, ,.i :...".-.:.:...... 4::.."::,::%.:;.. 1. ..:...... :.-: -- 1...... :...: :.,.. ".."zo..W..-.:N...... --.....1..:.....-; ....N ...&.-...... I..1 .:: ..,.' %s...... "-:I;...... -...... ::..:..,.:....,.,;F.'-!::. i:z,.,:..-s !W"... :V:: ... ..: ...... ,...... L...14 ..:...:...... 11.%:W,.....'.-...... ,X..: :... ".f."I...,....11 ...... -::-,,...,. ..,...... A,.",.,:::...,.. -.:4..:...",,%.... .4. i: ::!" ...... :O.%...... -ft.. ...X.. .Xii:.1, .."I'l ": .:::....:....-'..:...... 4... .- .5. ::,W.,: ::: s.. ..,...... -.:.:V:i::..?"W- ..' -..:I:. ::.,-.:..: 1,.1."","-,.: .: ....::.I...:. .11-:..-. V...... 1- -.. ,,--:,. .N..... --ff: .: ,:::..X.-:i...:.... .-.....:P.::.&..- %: :...... '..-.....-::..:. :.-...... :1...-44...:.. .-.. .:...... ::..-- ...... o.,:!,..:!:...... 04.:.,:1'e:.. :.:.:. ":As?:4- ..,...'..m...!;:1:....:4..4., 4. ,...... :-,....-f' :,...... ".:::: -:X4:. ... %....::.,%.q ..;,..1..1. ....-.:.,.L.-N .;! .;,?...W:::,...:...'..%%sV :-..: -1. ..-I ...- .....:. :.1:. :--....P...... :....H' .:.: ...... :.. ::1 .. ti i.... :; :.... -.-...... %.:: AIPO",,..."..-i.I:- %..:.-.-...-.:...... :.. '.. .::...%.::.. ..N...N:.:...:..-:..,..:...... - ...... :: ::N.: i.:.. N...%... :`,,I.:...... :.. ,.:.....I..%X;I -.. :iN:....4.;: .,-4...:i:A'...... ::-:;- ..1:.., :-.1 ...::%:.....::....: x:-:...... ,..: :.. A". .::::::.::.f:l.... .:: ,:::;.4-.Al...::.. ..:..:..:... ::. :`... N, 11.-...... -9.....: .4.....:.:.. -NN '.::,...:4.1..,...,.:,..:.--.,:::-:::.....:.:.i. "A.::!"::. .: .: ...... N". .:-...... '...-_:.-s- '.%V.... :-.:.!%::.-.. :::..4 .N.:%...... ::N.... '.....:...... :11.-liw: ..: -X.44'. :.-...... 4....-:.N 11. .-:I.;.:...... :...... ".:.....1- ..3.....:...:...:,_. :,....-,, W.:..O..:. ....N .....,....., '....:. ..._,i`AI-, "'.1 4:.. .:;:....-4-:-.,!4:: +...,..:N..1. 1: :::.,...... -..."r ..:..tlZ---.. O1:, .5.11-1I /l.....4'. a ..:: ,.:..1...: ..:.. .. .'. ,..4,:....IV ....."11.-.:!,:....::...-....:.:- ,...4. W.2:1 ;.:.:..':....iiF4u-.I....1::. .L .:, I1. :-.....s I...... :: :1 '': "_. :..1, Wil:...:....N.-.....i?.:-j;`-kM.!::.-^. -.-.&..&Is.":...1. Protected by copyright. :X -.....I http://jnnp.bmj.com/ on September 29, 2021 by guest. J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.51.3.332 on 1 March 1988. Downloaded from

CT-assisted stereotactic brain biopsy: value of intraoperative frozen section diagnosis 341 very important in considering the grade of malig- 11 Bullard DE, Osborne D, Burger PC, Nashold BS Jr. Further nancy. In this respect, the pathologist must be famil- experience utilizing the Gildenberg technique for computed iar with the clinical spectrum of neurological disease tomography-guided stereotactic biopsies. Neurosurgery 1986;19:386-91. and must be aware of the fact that a knowledge of 12 Gildenberg PL, Kaufman HH, Murthy KSK. Calculation of CT-histology correlations by the neurosurgeon may stereotactic coordinates from the computed tomographic scan. enhance the value of the biopsy procedure. Neurosurgery 1982;10:580-6. Although frozen section preparations have been 13 Heilbrun MP, Roberts TS, Apuzzo MLJ, Wells TH, Sabshin JK. accused of frequently obscuring the fine cellular Preliminary experience with Brown-Roberts-Wells (BRW) computerized tomography stereotaxic guidance system. detail,2' an accurate judgement of the type of lesion J Neurosurg 1983;59:217-22. and often the grade of malignancy can generally be 14 Kelly PJ, Earnest F IV, Kall BA, Goerss SJ, Scheithauer B. made on frozen section examination alone. It is clear, Surgical options for patients with deep-seated brain tumors: computer-assisted stereotactic biopsy. Mayo Clin Proc however, that the combination of CT-guided stereo- 1985;60:223-9. tactic techniques and the meticulous application of 15 Ostertag CB, Mennel HD, Kiessling M. Stereotactic biopsy of the frozen section technique is responsible for this brain tumors. Surg Neurol 1980;14:275-83. high degree of accuracy. 16 Daumas-Duport C, Monsaingeon V, N'Guyen JP, Missir 0, Szikla G. Some correlations between histological and CT aspects of cerebral gliomas contributing to the choice of References significant trajectories for stereotactic biopsies. Acta Neurochir [Suppl] 1984;33:185-94. I Araki T, Inouye T, Suzuki H, Machida T, lio M. Magnetic 17 Hitchon PW, Schelper RL, Barloon T. Accuracy of CT scans in resonance imaging of brain tumors: Measurement of TI. identifying tumor tissue. Appl Neurophysiol 1985;48:463-6. Radiology 1984;150:95-8. 18 Scerrati M, Rossi GF. The reliability of stereotactic biopsy. Acta 2 Brant-Zawadzki M, Badami JP, Mills CM, Norman D, Neurochir [SupplJ 1984;33:201-5. Newton TH. Preliminary intracranial imaging: a comparison 19 Kleihues P, Volk B, Anagnostopoulos J, Kiessling M. of magnetic resonance and CT. Radiology 1984;150:435-40. Morphologic evaluation of stereotactic brain tumour biopsies. 3 Kendall BE, Jakubowski J, Pullicino P, Symon L. Difficulties in Acta Neurochir [SupplJ 1984;33:171-81. diagnosis of supratentorial gliomas by CAT scan. J Neurol 20 Badt B. Mikroskopische Schnelldiagnose bei hirnchirurgischen Protected by copyright. Neurosurg Psychiatry 1979;42:485-92. Eingriffen. Zbl Neurochir 1937;2:123-39. 4 Latchaw RE, Gold LHA, Moore JS Jr, Payne JT. The non- 21 Burger PC. Use of cytological preparations in the frozen specificity of section absorption coefficients in the differentiation of diagnosis of central nervous system neoplasia. Am J Surg solid tumors and cystic lesions. Radiology 1977;125:141-4. Pathol 1985;9:344-54. 5 Latchaw RE, Payne JT, Loewenson RB. Predicting brain tumor 22 Eisenhardt L, Cushing H. Diagnosis of intracranial tumors by histology: change of effective atomic number with contrast supravital technique. Am J Pathol enhancement. AJNR 1930;6:541-52. 1980;1:289-94. 23 McMenemey WH. An appraisal of smear-diagnosis in neu- 6 Lee BCP, Kneeland JB, Cahill PT, Deck MDF. MR recognition rosurgery. Am J Clin Pathol 1960;33:471-9. of supratentorial tumors. AJNR 1985;6:871-8. 24 Marshall LF, Adams Graham DI. The 7 H, Doyle D, histological Spiegel EA, Wycis HT, Marks M, Lee AJ. Stereotaxic apparatus accuracy of the smear technique for neurosurgical biopsies. J for operations on human brain. Science 1947;106:349-50. Neurosurg 1973;39:82-8. 8 Apuzzo MLJ, Sabshin JK. Computed tomographic guidance 25 Morris AA. The use of the smear technique in the rapid histolog- stereotaxis in the management of intracranial mass lesions. ical diagnosis of tumors of the central nervous system: descrip- Neurosurgery 1983;12:277-85. tion of a new staining method. J Neurosurg 1947;4:497-504. 9 Bakay RAE, Hoffman JC, Rhodes M. Advantages of computer- 26 Russell DS, Krayenbuhl H, Cairns H. The wet film in ized technique tomography-assisted interaction during stereotactic pro- the histological diagnosis of intracranial tumours; a rapid cedures using computer-generated graphics. Appl Neurophysiol method. J Path Bact 1937;45:501-5. 1985;48:79-80. 27 Takei Y. Pathology of pituitary tumors and value of frozen 10 Boethius J, Collins VP, Edner Lewander J. G, R, Zajicek Stereo- section diagnosis. In: Tindall GT, Collins WF, eds. Clinical http://jnnp.bmj.com/ tactic biopsies and computer tomography in gliomas. Acta Management of Pituitary Disorders. New York: Raven Press, Neurochir 1978;40:223-32. 1979:93-112. on September 29, 2021 by guest.