Functional MRI and Intraoperative Brain Mapping to Evaluate Brain Plasticity in Patients with Brain Tumours and Hemiparesis

Home , Franciscus Sylvius

J Neurol Neurosurg Psychiatry 2000;69:453–463 453 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.69.4.453 on 1 October 2000. Downloaded from Functional MRI and intraoperative brain mapping to evaluate brain plasticity in patients with brain tumours and hemiparesis

F E Roux, K Boulanouar, D Ibarrola, M Tremoulet, F Chollet, I Berry

Abstract Since the works of Penfield and Boldrey1 and Objective—To support the hypothesis Foerster2 intraoperative cortical stimulation to about the potential compensatory role of precisely map brain functions has proved to be ipsilateral corticofugal pathways when the useful, especially in infiltrative tumours in or 34 contralateral pathways are impaired by near eloquent areas. Using optical imaging, 5 brain tumours. Haglund et al demonstrated the reliability and Methods—Retrospective analysis was car- the accuracy of this method to map function ried out on the results of functional MRI precisely. This technique has become a stand- (fMRI) of a selected group of five paretic ard method for localising language and sensory motor cortex in patients during patients with Rolandic brain tumours who neurosurgical procedures.67 Use of this tech- exhibited an abnormally high ipsilateral/ nique is, however, restricted to patients who contralateral ratio of activation—that is, undergo open brain surgery. In paretic pa- movements of the paretic hand activated tients, the existence of completely negative predominately the ipsilateral cortex. stimulations of the hand area normally inner- Brain activation was achieved with a flex- vating the paretic limb,7 and possibly related to ion extension of the fingers. Statistical plasticity phenomena,8 has seldom been docu- parametric activation was obtained using mented by functional studies. In patients with a t test and a threshold of p<0.001. These brain tumours, it has also been shown that patients, candidates for tumour resection, functional MRI (fMRI) with blood oxygen also underwent cortical intraoperative dependent (BOLD) contrast was capable of stimulation that was correlated to the depicting functional areas,9–11 providing func- fMRI spatial data using three dimensional tional information complementary to the reconstructions of the brain. Three pa- structural studies. Thus, fMRI has been used tients also had postoperative control in patients with brain tumours for surgical fMRI. planning,10 12 13 epilepsy surgery,12 13 and to Results—The absence of fMRI activation detect functional reorganisations resulting 11 of the primary sensorimotor cortex nor- from structural or functional damage. Re- mally innervating the paretic hand for the cently, fMRI in paretic patients with brain threshold chosen, was correlated with tumours has shown an abnormally high completely negative cortical responses of ipsilateral/controlateral ratio of activation http://jnnp.bmj.com/ when the subject performs a task with his the cortical hand area during the opera- aVected hand—that is, the cortex ipsilateral to tion. The preoperative fMRI activation of INSERM 455, Hôpital the paretic hand activates.14 Caramia et al,15 PURPAN, F-31059 these patients predominantly found in the using motor evoked potentials (MEPs), sug- Toulouse, France ipsilateral frontal and primary sensori- gested that ipsilateral activation could be sup- F E Roux motor cortices could be related to the K Boulanouar pressed or undetected in the normal brain but F Chollet residual ipsilateral hand function. could be detected when the contralateral con-

I Berry Postoperatively, the fMRI activation re- trol becomes impaired by a tumour. However, on October 1, 2021 by guest. Protected copyright. turned to more classic patterns of activa- the persistence of movement in the aVected Department of tion, reflecting the consequences of hand could also be related to persistant Neurosurgery therapy. F E Roux contralateral control that is undetected by the M Tremoulet Conclusion—In paretic patients with functional studies. brain tumours, ipsilateral control could be Combining fMRI and cortical stimulation, Department of implicated in the residual hand function, we tried to support the hypothesis of a Neuroradiology when the normal primary pathways are potentially compensatory role of ipsilateral D Ibarrola impaired. The possibility that functional control when contralateral routes are ineVec- I Berry tissue still remains in the peritumorous tive. For this purpose, we studied five paretic Department of sensorimotor cortex even when the preop- patients in whom the lack of fMRI activation Neurology erative fMRI and the cortical intraopera- seen in the contralateral primary sensorimotor F Chollet tive stimulations are negative, should be region of the aVected hand was correlated with the absence of response with cortical stimula- Correspondence to: taken into account when planning the Dr Franck-Emmanuel Roux tumour resection and during the operation. [email protected] tion. (J Neurol Neurosurg Psychiatry 2000;69:453–463) Materials and methods Received 26 August 1999 PATIENTS and in final form 5 May 2000 Keywords: brain tumour; functional MRI; brain plastic- Five patients (two men; three women; age Accepted 5 May 2000 ity; cortical stimulation range 50 to 68 years; median age-60 years; all

www.jnnp.com 454 Roux, Boulanouar, Ibarrola, et al J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.69.4.453 on 1 October 2000. Downloaded from

right handed) presenting with various tumours gyrus (Brodmann area 4), and sometimes the near or in the motor strip and candidates for a supplementary motor area.9 We chose an fMRI tumour resection were studied. These patients task close to the movement elicited intraopera- were selected from a larger series of 32 patients tively by stimulation (flexion of the fingers). with various Rolandic tumours who underwent Furthermore, the hand area is functionally motor fMRI and cortical and subcortical important, usually well represented in the stimulations at our institution in the past 2 motor cortex, and easy to test by direct stimu- years. These five patients were selected because lation. Although the task chosen was easy to all of them had a particular pattern of fMRI perform, all subjects were trained to rehearse activation: making a movement with their the task a few minutes before the procedure. paretic hand, they activated predominantly or Patients were instructed to do the task as fast as exclusively the ipsilateral hemisphere (abnor- possible, but without (or with minimal) head mally high ipsilateral/contralateral ratio with an motion. increased activity of the ipsilateral hemisphere). Emergence of a rapidly progressive fMRI data acquisition motor weakness was the presenting symptom Patients were positioned in the head coil of a in these five patients, but epilepsy was also 1.5T Magnetom Vision® MR scanner (Sie- noted in one patient. We used the Canadian mens, Erlangen, Germany). Optimisation of neurological score16 to evaluate globally the the magnetic field was performed with the degree of hemiparesis of each patient. More automatic map-shim procedure to reach a specifically, the motor impairment of the hands gradient tolerance of 0.001mT/m. fMRI data in these patients was also assessed by a finger were obtained using a GE-EPI single shot ballistic opposition task and by the Medical sequence (TE=60 ms; FA=90°; slice Research Council (MRC) scale.17 The time number=10, matrix size=64×64, FOV=200 required to perform 20 finger oppositions of mm, slice thickness=5 mm, distant factor=0.5 each hand was recorded and compared. The mm). The 10 slices were positioned parallel to degree of motor impairment could thus be the anterior commissure and the posterior assessed by the time comparison between both commissure (AC/PC) axis from the base of the hands. The MRC scale is based on an estima- brain to the vertex. A staV member was always tion of the index abduction power, as a gross present near the patient during the acquisitions evaluation of hand motor skill.17 The duration to control the procedure, to encourage the of the presenting symptom before the diagnosis patient to do the task to the best of their ability, was made at the time of surgical intervention and to ensure that the patient followed the start ranged from 3 days to 2 weeks. and stop signals. In fact, the fMRI procedure Patients were also assessed between 6 weeks can be long for this category of patient with and 10 weeks after the operation with the hemiparesis and often with high grade tu- Canadian neurological and the MRC scores, mours. The presence of a staV member and the finger opposition test. Their degree of throughout the MRI study near the patient was postoperative impairment or recovery was also useful to ensure that the patient had no defined as changes in performance in these visible syncinesis of his normal hand when per- tests. Hand motor recovery was arbitrarily forming the task with the aVected hand. defined as an increase of at least 1 point in the During the procedure, the patient alternated

MRC score and an improvement of the finger epochs of rest and epochs of activation. Each http://jnnp.bmj.com/ opposition time of the paretic hand versus nor- epoch (rest or activation) lasted 30 seconds mal hand. Similarly, postoperative motor im- while 10 images were acquired every 3 seconds. pairment was defined as a reduction of at least Alternative rest and activation periods were 1 point of the MRC score and by a deteriora- repeated four times; with each the procedure tion of the finger opposition time of the paretic began with a period of rest. Each period was hand versus normal hand. controlled vocally by the headphones. Four patients also had a control fMRI

FMRI PROCEDURE procedure after their operation with the same on October 1, 2021 by guest. Protected copyright. Motor task procedure task. But because of head motion, one study The task chosen was a flexion and extension of has to be discarded. Because these control the fingers of the paretic hand. The good coop- fMRI studies were done only for research pur- eration of the patients during the fMRI proce- poses, no three dimensional anatomical run dure allowed us to also test the normal hand. was done postoperatively. We considered that it This specific task was chosen because of its was more acceptable for the patients not to ability to activate most precisely the precentral again undergo a 12 minute three dimensional anatomical sequence. Thus, for the postopera- Table 1 Summary of the global assessment of the patients, fMRI performed, clinical evaluations, and intraoperative mapping procedures tive studies, the echo planar images were not realigned on the anatomical ones. All these Preoperative data are summarised in table 1. fMRI Intraoperative cortical Full Full preoperative mapping and postoperative Postoperative PH NH motor assessment electrocorticography motor assessment fMRI fMRI data analysis The fMRI data were analysed with the Statisti- 1+ Dis+ + + + 18 2+ + + + + NP cal Parametric Map (SPM) 96 software 3+ + + + + + (Wellcome Department of Cognitive Neurol- 4+ + + + + + ogy, London, UK) and performed on a Sun 5+ + + + + Dis SPARC workstation (Mountain view, CA, NP=Not performed; PH=paretic hand; NH=normal hand; Dis=discarded (head motion). USA). The first three images of each run were

www.jnnp.com Brain plasticity in patients with brain tumours 455 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.69.4.453 on 1 October 2000. Downloaded from

discarded to allow signal stabilisation and The anatomical images were analysed with the remaining 77 volumes of 10 slices three dimensional views of the surface of the were realigned to correct the subject brain. The central sulcus was found using the movement during scanning using the three dimensional reconstructions of the ﬁrst volume of images as reference. Then surface of the brain. Intraoperative cortical

Postoperative time for 20 finger oppositions the detection of the activated voxels was stimulation was used to localise the hand area performed on a pixel by pixel basis. If and eventually other areas of functional cortex large movements of the head occurred, in the Rolandic cortex after determination of the data were excluded from analysis. We the afterdischarge threshold. The cortex was used the general linear model imple- directly stimulated using a bipolar Ojemann mented in SPM 96 where conditions cortical stimulator (1 mm electrodes separated (rest or activation) stand for independant by 5 mm:Radionics®, Burlington, USA), with Preoperative time for 20 finger oppositions variables after global normalisation to an appropriate current to try to elicit a specific cancel diVerences among scans. Usually, movement (flexion of the fingers). The current voxels were considered as significant if amplitude was progressively increased by 1 p<0.001, corrected for multiple com- mA beginning at 1 mA. We used a standard parison. In each patient, four regions procedure of stimulation with biphasic square were studied using reliable sulcal land- wave pulses of 1 ms at 60 Hz, with a maximum

Postoperative MRC score marks. The regions studied bilaterally train duration of 4 seconds. Electrocorticogra- were the primary sensorimotor (S1/M1), phy was used to determine the afterdischarge frontal (BA 8; BA 9), temporoparietal threshold. When a functional site was found, it (BA 40; BA 7; BA 22; BA 39) regions, was marked by a sterile ticket of 0.25 cm2 and and supplementary motor regions ( BA then another area 5 mm away was tested. The 6) and the cingular gyrus (BA 24; BA exact locations of functional sites were com- PH NH PH NH PH NH PH NH Preoperative MRC score 32). The images were reformatted to be pared with the three dimensional reconstruc- integrated in the radiological atlas of tions of the brain surface and marked on these Talairach and Tournoux19 using the reconstructions. We studied the whole area ANALYZE software (Mayo Clinic, Ro- exposed during craniotomy by cortical and chester, MN, USA). The results obtained subcortical mapping as well. Our policy was to Postoperative Canadian neurological score in both hemispheres were then compared spare the functional areas found by this test and correlated with cortical brain map- during the tumour removal by resecting the ping. tumorous tissue no more than 1 cm from eloquent cortex (distance of the resection margin INTRAOPERATIVE CORTICAL STIMULATION from the nearest functional site). When no Preoperative Canadian neurological score PROCEDURE AND CORRELATION functional site was found, especially in the Three dimensional data acquisition presumed hand area, our policy was to spare Local distortions of the surface of the the precentral and postcentral gyri. Intraop- gyri produced by the tumour were found erative photographs of the brain were taken

Time between the onset of symptoms and fMRI on three dimensional reconstructions. with the sites of positive or negative cortical This allowed preoperative and intraop- stimulation. The data were analysed by visual erative computer assisted assessment of comparison of both surface renderings and the relations between the tumour and the views of the operative ﬁelds. The exact

important landmarks of gyral anatomy. locations of the functional areas were inte- http://jnnp.bmj.com/ The surfaces of the brain were recon- grated on the three dimensional brain surface structed from a three dimensional data renderings by using anatomical landmarks on set obtained with a 3D-MPRAGE se- the brain’s surface (gyral and sulcal pattern). quence (TR=15 ms; TE=7 ms; FA=12°; Because of the shape and the location of the 128 partitions; FOV=300 mm; matrix craniotomy, direct stimulation of some areas size=256×256, slab thickness=150 mm, (especially the supplementary motor area and 3 acial seizures left hand hemiparesis 13 days 8voxel size=1.17 9×1.17 2×1.17 5 mm , 3NA=1, 5cingular imp. 11 gyrus) 52 was 12 not always possible,

TA=10 minutes). The centre of the three making the validation by direct stimulation of on October 1, 2021 by guest. Protected copyright. dimensional block was positioned ac- these areas diYcult. Clinical features cording to the AC/PC plane. These three dimensional sequences lasted 12 minutes Results ® each. We used the ANALYZE software CLINICAL FINDINGS (Mayo clinic, Rochester, MN, USA) to Clinical presentations of our patients at the separate the brain from the overlying onset of their hemiparesis as well as at the time skull and scalp.The 120 slices of the three of their postoperative assessment are summa- dimensional data set were usually edited rised in table 2. The MRC and finger in 90 to 120 minutes. opposition scores showed that all our patients were severely impaired before the operation. Cortical stimulation procedure Two of them (patients 1 and 3) were unable to Patients were intubated and operated on perform the finger opposition task. under general anaesthesia using propofol infusion and 50% nitrous oxide in FMRI RESULTS oxygen with supplemental 1% isoflurane From 15 fMRI studies (10 preoperative and and fentanyl (without muscle relaxant). five postoperative), two studies had to be A large craniotomy was often required to discarded because of excess motion (the study ensure that the whole Rolandic region of the normal hand of patient 1 and the control 1234 M/505 F/62 F/58 Glioblastoma, R hemisphere M/68 Astrocytoma grade III, F/63 R hemisphere Meningioma, L hemisphere Lung metastasis, R hemisphere Bapidly progressive hemiparesis Astrocytoma grade III, Brachiof R hemisphere Left hand hemiparesis Left hand Brachiofacial hemiparesis hemiparesis 1 days 7 4 days 5 days 7 days 9.5 7.5 7.5 9 3 10 9 9 5 3 3 5 4 5 5 5 5 5 4 59 4 5 5 15 5 imp. 44 24 18 38 11 17 10 21 28 17 28 10 12 Patient Sex/age Pathological localisation Imp=Task impossible to perform. Only flexion and extension of the fingers possible. Table 2 Summary of the clinical assessment of the patients could be studied by cortical stimulation. study of patient 5. In patient 2, the control

www.jnnp.com 456 Roux, Boulanouar, Ibarrola, et al J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.69.4.453 on 1 October 2000. Downloaded from

Table 3 Number of signiﬁcantly (p<0.001) activated voxels in the paretic hand of each patient before and after operation

Primary sensorimotor Supplementary motor area (S1, M1) Frontal regions Temporoparietal regions area and cingular gyrus

Hemisphere Before op After op Before op After op Before op After op Before op After op

1 Ipsilateral 89 0 140 29 38 0 147 0 Contralateral 0 114 0 14 0 0 90 0 2 Ipsilateral 110 NP 48 NP 11 NP 11 NP Contralateral 0 NP 13 NP 0 NP 16 NP 3 Ipsilateral 47 6 30 8 5 0 5 5 Contralateral 0 34 12 0 0 0 0 11 4 Ipsilateral 87 0 32 9 10 4 2 25 Contralateral 0 31 14 4 0 16 2 19 5 Ipsilateral 56 Dis 17 Dis 9 Dis 19 Dis Contralateral 0 Dis 0 Dis 0 Dis 12 Dis

Op=Operation; NP=not performed; Dis=discarded.

study of the paretic hand was also not diVerent from the ratio found when the performed because we decided not to do the contralateral, paretic hand was tested. These fMRI procedure, the patient being too tired data are summarised in table 4. after the anatomical control MRI runs. No significant foci of increased activation were found within the tumours. Control fMRI Four patients (patients 1, 3, 4, and 5) had an fMRI after their operation. In patient 5, the Paretic hand images had to be discarded because of In this group of patients the most common movement artifacts. For these control fMRI regions with increased activation were in the only the paretic hand was tested. In patient 1, unaVected hemisphere, especially in the pri- harbouring a right hemispheric glioblastoma mary sensorimotor areas, but also in the frontal and who had a partial recovery of his hand, the and medial areas. Ipsilateral activation was not control fMRI done 2 months after the exclusive and contralateral activation to the operation showed the return of a classic paretic hand was also seen individually and to a contralateral primary sensorimotor activation lesser degree, especially in frontal and medial with no ipsilateral activation (figs 1 and 2). areas. A strong ipsilateral/contralateral ratio in These findings were probably related to motor the sensorimotor area has been found in each recovery of the patient and to elimination of the patient. In four patients (patients 1, 3, 4, and mass eVect. Patient 3 presenting with a menin- 5), we found no controlateral primary sensori- gioma underwent a control fMRI 6 weeks after motor area activation to the paretic hand. her operation that also showed a contralateral Patient 2 had only four activated pixels in his primary sensorimotor activation to the former primary sensorimotor area contralateral to the paretic hand. In patient 4, fMRI showed a sig- paretic hand. These data are summarised in nificant change in the location of the activa- table 3. tions when compared with the previous ones.

This fMRI has been done 8 weeks postopera- http://jnnp.bmj.com/ tively. A contralateral activation in the primary Normal hand sensorimotor area was again visible with a less The activations were found predominantly in important ipsilateral/contralateral ratio. the contralateral, unaVected hemisphere and were in agreement with the classic pattern of hand activation. They were located principally Cortical stimulation results and fMRI correlation in the contralateral primary sensorimotor area The whole region of the craniotomy has been

(mean of the activated pixels in the primary studied carefully by bipolar stimulation, repeat- on October 1, 2021 by guest. Protected copyright. sensorimotor 227). In three patients (patients ing the stimulations at least twice for each site. 2, 3, and 4), no activated pixels were found in In each patient we progressively increased the the ipsilateral primary sensorimotor area (af- level of the stimulation intensity because the fected hemisphere). All patients had a high amount of current that elicits a hand move- contralateral/ipsilateral ratio, individually very ment is variable individually and cannot be predicted. In our patients no movement of the Table 4 Number of significantly (p<0.001) activated voxels in the normal hand of each patient before the operation aVected hand was recorded even at the highest level of stimulation (16 mA) but in one patient Primary Supplementary this caused a brachiofacial seizure that resolved sensorimotor Frontal Temporoparietal motor area and with application of cold Ringer’s lactate to the Hemisphere regions (M1/S1) egions regions cingular gyrus cortex. It is possible to find responses elicited at 1 Ipsilateral 5 0 0 8 higher stimulus intensities, but this is not with- Contralateral 230 45 0 0 2 Ipsilateral 0 0 0 17 out risk and we tried to avoid any suprathresh- Contralateral 147 13 0 7 old stimuli because it can result in current 3 Ipsilateral 0 6 0 12 spread, neural fatigue, or cell habituation. In Contralateral 143 0 19 0 4 Ipsilateral 0 0 0 7 this group of patients, we were unable to elicit Contralateral 192 0 0 5 any hand movement by cortical stimulation. If 5 Ipsilateral 3 4 0 0 no hand movement was found, direct cortical Contralateral 98 30 0 11 stimulation produced flexion of the elbow

www.jnnp.com Brain plasticity in patients with brain tumours 457 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.69.4.453 on 1 October 2000. Downloaded from

Figure 1 Patient 1. A 50 year old patient presenting with a right glioblastoma and severe left hemiparesis. He was unable to do the finger ballistic opposition task but flexion and extension of the fingers was still possible. Preoperative fMRI done 13 days after the onset of the symptoms (brachiofacial seizures and rapidly progressive hemiparesis) realigned on the anatomical slices (significant threshold: p<0.001). Paretic hand studied with fMRI during flexion and extension of the fingers. The activations were localised in the ipsilateral sensorimotor cortex and in the supplementary motor areas. No activation in the controlateral primary sensorimotor area can be seen. This finding was confirmed by the cortical brain mapping done intraoperatively.

Figure 2 The same patient as in fig 1 and the same task but with fMRI done 2 months after the operation. The patient had a partial recovery of his left hand (finger opposition task done slowly—52 seconds v 12 seconds for the normal hand). Left hand studied with fMRI during flexion and extension of the fingers (significant threshold: p<0.001). This fMRI study was done only for research purposes, and for the comfort of the patient (the three dimensional anatomical sequence lasting 12 minutes), no three dimensional anatomical run was done postoperatively; the echo planar images have thus not been realigned on the anatomical ones. The pattern of activation is now more classic with an activation of the controlateral sensorimotor area while the patient is performing the task. No activation can be seen in the ipsilateral cortex. Some artefacts due to the operation are visible on the periphery of the operated hemisphere. (n=1), abduction of the shoulder (n=3), and rapidly seen after the onset of symptoms (a few contraction of the controlateral face (n=2) (figs days), could be temporary, and that an absence 3 and 4). This last movement was often of detection of a functional tissue by fMRI or diYcult to see because of the operative position cortical stimulation did not imply that viable

of the patient (head turned on the head rest, functional tissue was present in the peritumor- http://jnnp.bmj.com/ masked by the drapes). We found no move- ous brain. Our results are in agreement with ment in the lower limbs, and no attempt was those of Yoshiura et al, who used fMRI in seven made to expose the leg cortical areas. patients with brain tumours and found that the Thus, the absence of fMRI activation in the ratio of the ipsilateral/contralateral activated primary sensorimotor area of the aVected hand was abnormally high in three paretic hemisphere was associated, in each patient, patients.14 Caramania et al used MEPs in seven with a negative cortical stimulation (impossible patients with various brain tumours and in 15

to elicit any hand movement even with high volunteers. Not all their patients had motor on October 1, 2021 by guest. Protected copyright. intensities). Postoperatively, partial or total deficits but they found, comparing the results recovery was related to significant changes in with healthy subjects, that ipsilateral MEPs the fMRI studies (figs 5, 6, 7, and 8). were generally absent in normal subjects but present in the patients by stimulating the non- aVected hemisphere. These findings could be Discussion related to a potential compensatory role of the Combining preoperative and postoperative ipsilateral motor control. fMRI and cortical stimulation in a selected group of patients with hemiparesis and brain tumour, we demonstrated (a) that the lack of METHODOLOGICAL ISSUES fMRI activation seen in the contralateral They are many methodological issues related primary sensorimotor area of the aVected hand to the use of fMRI in patients with brain was correlated with the absence of response tumours. Technical problems such as echopla- with cortical stimulation; (b) that this fact pre- nar distortion, paradigm choices, movement sumes that the hand area in the aVected hemi- artifacts, and venous eVects have been evoked sphere was not functional or masked; (c) that in many fMRI studies.9 20–22 But more specific the residual hand function was probably problems must be recognised when using the related to the ipsilateral activation seen with fMRI technique in patients with brain tu- fMRI; (d) and that these phenomena could be mours. It is in these particular conditions that

www.jnnp.com 458 Roux, Boulanouar, Ibarrola, et al J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.69.4.453 on 1 October 2000. Downloaded from

Figure 3 Illustrative case of the spatial distribution of the ipsilateral activated area in the intact hemisphere. Patient 2 presenting with an astrocytoma grade III (WHO classification). She had fMRI 24 hours after the onset of her symptoms (a rapidly progressive left hemiparesis). The left paretic hand (below) and right normal hand (top) studied with fMRI during flexion and extension of the fingers (significant threshold: p<0.001). Functional images realigned on the anatomical ones. The recruited ipsilateral cortical area and the activated area of the normal hand in the intact hemisphere are in the same region. Most of the ipsilateral activation is located in the Rolandic zone and there is a certain degree of overlapping between both areas.

the validation of fMRI data is the most useful. also related to the specialisation of the cerebral The most appropriate statistical methodology structure studied.23. These factors can make in the analysis of fMRI data in neurosurgical the analysis of fMRI data extremely diYcult. patients has not yet been defined. The rate and Therefore relations between neuronal activity, the amplitude of the paradigm vary and local cerebral blow flow, and cerebral tumour influence21 the amount and volume of brain (especially in high grade astrocytoma) in fMRI activation obtained. The area of activation is remain unclear.23 The significance thresholds chosen to generate activation maps of a function are arbitrary and it is precisely the discrepancy between this statistical map and http://jnnp.bmj.com/ the true map of a function that requires validation. It could be argued that the arbitrary threshold chosen was too strict and that there might have been activations at a lower level of significance. The usefulness of intraoperative cortical mapping is that it can validate the statistical maps given by the analysis of fMRI. It was particularly true in our patients where the on October 1, 2021 by guest. Protected copyright. absence of activation in the primary sensorimotor area of the aVected hemisphere has to be validated by another technique.

NEUROANATOMICAL ASPECTS Türck first described in 1851 the pyramidal tract, extending from the motor cortex to the spinal cord.24 In 1909, Holmes and May showed that the pyramidal tract arose from the precentral gyrus.25 But others have shown that the pyramidal fibres originate from various cortical areas, including the precentral gyrus, and also the premotor cortex, the primary sensory cortex, the parietal associative cortex, and Figure 4 Patient 2. Intraoperative view of the cortex after brain mapping and before the supplementary area.26 At the junction of the tumour removal. Bipolar stimulation did not identify sites that elicited a flexion of the fingers but abduction of the controlateral shoulder was seen (A) as well as flexion of the medulla and the spinal cord, most of the fibres elbow (B). Negative sites for a motor response are marked N. Arrows=the central sulcus. cross the midline in the pyramidal decussation

www.jnnp.com Brain plasticity in patients with brain tumours 459 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.69.4.453 on 1 October 2000. Downloaded from

Figure 5 Patient 4. Paretic left hand studied with fMRI during flexion and extension of the fingers (significant threshold: p<0.001). The patient presenting with a lung metastasis in the right hemisphere. Most of the activation is situated in the ipsilateral hemisphere, in the Rolandic area.

Figure 6 Patient 4. Normal right hand activated area. fMRI during flexion and extension of the fingers (significant threshold: p<0.001). Classic pattern of activation with the Rolandic area of the intact hemisphere predominantly activated. to form the lateral corticospinal tract whereas adjoining portions of the intermediate zone of uncrossed fibres form the ventral corticospinal the grey matter of the spinal cord. It originates tract or remain in the lateral corticospinal principally in Brodmann’s area 6 and in the tract.26 Ipsilateral pyramidal fibres seems to Brodmann’s area 4 in zones controlling the represent 10% to 25% of the total fibres,26 neck and trunk.26 The anatomical existence of although quantitative assessment is diYcult an ipsilateral pyramidal tract can support the and probably varies individually. The ventral hypothesis of the partial contribution in the corticospinal tract projects bilaterally to the residual movement in paretic patients or parti- ventromedial motor neuron pools innervating pation in recovery from stroke or tumours,27 axial and proximal muscles as well as to the but this hypothesis must be supported by functional studies. http://jnnp.bmj.com/

INTRAOPERATIVE CORTICAL STIMULATION, fMRI CORRELATION, AND CLINICAL OUTCOME Data from fMRI has been recently validated prospectively by SEPs,20 electric transcranial stimulations,28 magnetoencephalography,28 and 29 direct intraoperative stimulation often in a on October 1, 2021 by guest. Protected copyright. few patients and principally in motor areas. Although cortical mapping identifies only those eloquent brain areas that are on the surface of the brain, it is still considerated as the gold standard of brain mapping.671030 All authors have emphasised the good spatial correlation existing between fMRI and this technique. In the primary sensorimotor areas, several authors28–30 have shown that functional activated MRI areas were related to a positive electrostimulation response and that when no activation was found on fMRI, no response was found intraoperatively by electrostimulation. We also found in our patients, by the usual methods of electrostimulation, that the absence Figure 7 Patient 4. Intraoperative view showing the post-Rolandic tumour (arrow). of fMRI activation in a given area was Bipolar stimulation with 16 mA did not identify sites that elicited a flexion of the fingers. correlated with the absence of response. The whole area of craniotomy has been studied. N=Negative sites for hand motor response; thread=the central sulcus. These findings are correlated with the fMRI data showing no Some criticism could arise about the spatial activation in the controlateral Rolandic area of the paretic hand (fig 5). correlation between fMRI and cortical stimula-

www.jnnp.com 460 Roux, Boulanouar, Ibarrola, et al J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.69.4.453 on 1 October 2000. Downloaded from

Figure 8 Patient 4 postoperative fMRI of the left hand (former paretic hand). fMRI during flexion and extension of the fingers (significant threshold: p<0.001). fMRI done 8 weeks postoperatively. This fMRI study was done only for research purpose, and for the comfort of the patient (the three dimensional anatomical sequence lasting 12 minutes), no three dimensional anatomical run has been done postoperatively; the echo planar images were thus not realigned on the anatomical ones. Note the echo planar artefacts in the operated zone. Compare with the fMRI done preoperatively. It is the same patient, the same task, and the same procedure of analysis of the images but the tumour has been removed and the patient’s hand, although still impaired, is less paretic than preoperatively.

tion, especially how to know, intraoperatively, patients (patients 1, 3, and 4) where we were that the cortical area found by stimulation able to have a postoperative MRI. All of our matches spatially the fMRI activated areas. patients recovered postoperatively, although This point could be an issue in our patients. this recovery was partial in three of them The use of visual inspection of both three (patients 1, 4, and 5). dimensional surface renderings and operative If cortical reorganisation can depend on fields by using anatomical landmarks on the temporal factors, tumour location can also play brain’s surface to correlate our data has already a part. If no motor response is seen in the been used by others.29 31 32. This method has motor strip, that can mean that hand area is proved to be reliable and accurate29 although possibly displaced because of the tumour. This the use of a technique of functional image had already been noticed by Penfield and guided surgery in which motor fMRI imaging Boldrey.1 Seitz et al, studying six patients with is registered to an infrared based frameless low and high grade tumours, suggested that the stereotactic device should improve the tumour location was predictive of the direction localisation.33 of the functional shift and the degree of The second point is the possibility of finding functional compensation.8 The functional no motor response in the motor strip by elec- shift, either ventrally or dorsally in the function trical stimulation. This possibility has been of the maximal tumour growth, could explain reported in other series.73435The first explana- some negative cortical stimulation found.8 tion is that no motor function is present in this Other conditions of negative stimulation in- area, precluding any motor response. This is clude acute brain swelling and too small a probably the case in our paretic patients. Mass craniotomy to expose the motor strip.7

eVect and local invasion can lead to loss of http://jnnp.bmj.com/ function. The time frame in which a local neu- FUNCTIONAL ASPECTS ronal area becomes non-functioning or de- Ipsilateral movements have been found in stroyed is not well known. Skirboll and patients with cortical stimulation, although not Ojemann,34 using direct cortical stimulation, often.1 Cortical stimulation using MEPs in found that functional tissue can be present in humans has been associated with motor the boundaries of an invasive tumour without responses in ipsilateral fingers.37 Ipsilateral loss of function initially. In fMRI, several activation has been also reported with fMRI in 38 39 authors found that an activated area can be normal subjects. Compared with the cont- on October 1, 2021 by guest. Protected copyright. located in the peritumorous brain,91029 cor- rolateral side, the activation was smaller and roborating the findings of Skirboll and weaker.38 The ipsilateral primary sensorimotor Ojemann.34 The progressive nature of tu- cortex can also be significantly implicated in mours can explain the potential reorganisation complex movements.39 In paretic patients, care phenomena described by some authors. It must be taken in the analysis of fMRI data. seems that a critical level of neuronal infiltra- With only five patients, our series was limited tion or destruction must be reached for a defi- but the ipsilateral activation seen was clearly cit to be seen clinically.3 If cortical invasion can diVerent from the activations usually found in be a cause of loss of function, a mass eVect by normal subjects; firstly, because our patients disruption or stretching of the pyramidal tracts showed no controlateral activation of the can also be involved. In these circumstances, primary sensorimotor area and secondly be- where a tumour has been resected, there can cause the ipsilateral activations seen were more be rapid improvement in motor function (and important that the activations usually found in even full recovery) caused by the decompres- normal subjects. sive eVect of the surgery.836This implies that, Sparse data are available to show whether in some paretic patients, negative motor severe motor deficit reduces the probability of stimulation does not mean that no function is finding focal activity in the motor cortex and potentially present in the area tested. We what type of relation can be established found this phenomenon in three of our between the degree of motor deficit and fMRI

www.jnnp.com Brain plasticity in patients with brain tumours 461 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.69.4.453 on 1 October 2000. Downloaded from

activation. Previous functional studies report a region has been well described by Jacobs and widespread hand activity in paretic patients.40 Donoghue.52 Corticospinal neurons projecting Furthermore, activations of the ipsilateral cor- to proximal muscles in an overlap region may tex, of the supplementary motor area, and of be disinhibited53 54 explaining the shift of a cor- prefrontal areas in patients with motor deficits tical functional area in lesions of the pyramidal more often than in intact patients have been tract.846 This phenonenon can occur within reported by several authors.9274041 Atlas et al hours of a peripheral nerve lesion.55 The have hypothesised that it may be a quantifiable ipsilateral activation in paretic patients with diVerence on fMRI between the tumour bear- brain tumours could be due to the same ing cortex and the normal hemisphere that mecanism of desinhibition. could be related to the extent of the patient The spatial distribution of the ipsilateral deficit.9 In patients with cerebral vascular mal- activated area in the intact hemisphere must formations, Schlosser et al showed that gross also be discussed. Most of our ipsilateral neurological deficits were always present when activation was located in the Rolandic zone and the activated cortex was displaced to a new there was a certain degree of overlapping anatomical location.42 This displacement of between both areas. This finding has been functions has been described in some chronic underlined by others.37 44 56 Huttenlocher and neurological diseases potentially related to Raichelson found that the ipsilateral and cont- functional recovery.43 44 But it can also be found rolateral pyramidal tract originate from the in strokes27 45 and was recently described in same areas in rats that had undergone neonatal tumours.42 46 Displacement of function, in- hemispherectomy.56 Meagaki et al,44 using crease of the activated hand area, and ipsilat- MEPs in a case of unilateral extensive cortical eral activation seem to be the mecanisms dysplasia, found that the abductor pollicis decribed by most authors. However, care must brevis response of the paretic and controlateral be taken in the analysis of this last eVect side originated from the same motor cortex. because of the potential eVect of the handed- The fact that ipsilateral hand representation ness on the activations. Kim et al and Li et al, lies in the primary motor cortex was also using fMRI in volunteers, have reported a described by Wasseramn et al.37 degree of ipsilateral activation greater during As we demontrated previously in patients the performance of left hand tasks than during after stroke,27 57 our results suggest that the right hand tasks, especially in right handers.47 48 ipsilateral corticofugal pathways in paretic However, this phenomenon seems to occur to a patients are able to sustain not only proximal lesser degree with motor tasks than with but also distal residual activity without re- sensory tasks.48 Another problem is that we course to the controlateral primary motor cannot completely rule out associated move- areas. But this ipsilateral activation in paretic ment of the hand ipsilateral to the tumour patients with a brain tumour leads to some although we were physically present during the questions: are the findings explained in terms fMRI procedure to ensure that the patient have of reorganisation or are the paretic patients just no synkinesis of his normal hand. EMG trying harder to do the movement? It has been recordings were not made in our patients. Sub- shown that a certain degree of ipsilateral clinical associated movements of the normal activation can be seen in complex movement or hand have been noted in paretic patients.40 in patients doing some movement requiring a 58 Aberrant mapping of cortical function has more important force. A movement, becom- http://jnnp.bmj.com/ been well described in the developing brain.43 ing more complex or stronger, requires the The underlying mecanisms proposed to sustain participation of proximal muscles to stabilise brain plasticity are collateral axonal sprouting, the joints.These proximal muscles are more unmasking of pre-existing areas due to bilaterally represented in the cortex26 than dis- desinhibition,49 and contribution of the ipsilat- tal hand muscles and the ipsilateral activations eral hemisphere through the direct pyramidal could be partially explained by the recruitment pathway27 or via transcallosal connections. In of these more proximal muscles in a patient

our patients with acquired lesions, the ipsilat- trying to do the movement at best. This fact on October 1, 2021 by guest. Protected copyright. eral activations seen are probably related to the can also be supported by the EMG study of unmasking mecanism. Patient 2, for instance, Turton et al that in patients after stroke, ipsilat- had her fMRI 24 hours after the onset of hemi- eral responses from the unaVected hemisphere paresis. It would be unlikely that sprouting with were most prevalent in the proximal muscles of consequent formation of new synaptic contact the aVected limb.59 had occurred. The possibility of finding no function in the primary controlateral motor Conclusion cortex in patients able to sustain residual motor Ipsilateral activation has been described with activity and then to find an activated area in the fMRI in normal subjects. In patients with primary motor cortex on control fMRI after motor deficits, activations of the ipsilateral cor- the operation suggests a temporary disinhibi- tex, of the supplementary motor area, and of tion mechanism. Disinhibition of pre-existing prefrontal areas have also been reported more but normally weak or non-functional synaptic often than in intact patients.9274041 These connections, not ordinarily shown by standard displacements of functions seen in fMRI possi- electrophysiological assays, could be the sub- bly relate to brain plasticity phenomena42 and strate for short term changes occuring after need to be validated by other methods. In our interruption of the major motor pathways.50 51 patients and for the specific hand tasks used, The disinhibition mechanism leading to an the absence of primary sensorimotor activation increase of activation in an adjacent motor was correlated with negative cortical stimula-

www.jnnp.com 462 Roux, Boulanouar, Ibarrola, et al J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.69.4.453 on 1 October 2000. Downloaded from

tion. Disinhibition of pre-existing but normally tem. Vol II. Motor control. Bethesda: American Physiological Society, 1981;597–666. weak or non-functional synaptic connections 27 Cholet F, DiPiero V, Wise RJ, et al. The functional anatomy could be the mechanism implicated in the ipsi- of motor recovery after stroke in humans: a study with positron emission tomography. Ann Neurol 1991;29:63–71. lateral activations seen in our patients. How- 28 Morioka T, Mizushima A, Yamamoto T, et al. Functional ever, in paretic patients with Rolandic tu- mapping of the sensorimotor cortex: combined use of magnetoencephalography, functional MRI, and motor evoked mours, the possibility of finding no primary potentials. Neuroradiology 1995;37:526–30. sensorimotor activation in the peritumorous 29 Jack CR, Thompson RM, Butts RK, et al. Sensory motor cortex: correlation of presurgical mapping with functional brain would not imply that no functional tissue MR imaging and invasive cortical mapping. Radiology is potentially present even with a negative hand 1994;190:85–92. 30 Krings T, Buchbinder BR, Butler WE, et al. Functional cortical area stimulation. This should be taken magnetic resonance imaging and transcranial stimulation: into account in the presurgical evaluation of complementary approches in the evaluation of cortical motor function. Neurology 1997;48:1406–16. the functional tissue surrounding a tumour in 31 Yetkin FZ, Mueller WM, Morris G, et al. Functional MR paretic patients with fMRI. activation correlated with intraoperative cortical mapping. Am J Neuroradiol 1997;18:1311–15. 32 Fried I, Nenov VI, Ojemann SG, et al. Functional MR and 1 Penfield W, Boldrey E. somatic motor and sensory represen- PET imaging of rolandic and visual cortices for neurosur- tation in the cerebral cortex of man as studied by electrical gical planning. J Neurosurg 1995;83:854–61. stimulation. Brain 1937;60:389–443. 33 Schulder M, Maldijian JA, Liu WC. Functional image- 2 Foerster O. The motor cortex in man in the light of Hugh- guided surgery of intracranial tumors located in or near the lings Jackon’s doctrine. Brain 1936;59:135–59. sensorimotor cortex. J Neurosurg 1998;88:412–18. 3 Ojemann JG, Miller JW, Silbergeld DL. Preserved function 34 Gregory EM, Goldring S. Localization of function in the in brain invaded by tumor. Neurosurgery 1996;39:253–9. excision of lesions from the sensorimotor cortex. J 4 Skirboll SS, Ojemann GA, Berger MS, et al. Functional cor- Neurosurg 1984;61:1047–54. tex and subcortical white matter located within gliomas. 35 King RB, Schell GR. Cortical localization and monitoring Neurosurgery 1996;38:678–85. during cerebral operations. J Neurosurg 1987;67:210–19. 5 Haglund MM, Ojemann GA, Blasdel GG. Optical imaging 36 Lüders HO, Comair YG, Bleasel AF, et al. Recovery of of bipolar cortical stimulation. J Neurosurg 1993;78:785– function following lesions of eloquent brain areas. In: Fre- 93. und HJ, Sabel BA, Witte OW, eds. Brain plasticity, advance 6 Berger MS, Kincaid J, Ojemann GA, et al. Brain mapping in neurology. Philadelphia: Lippincott-Raven, 1997;73:335– techniques to maximize resestion, safety and seizure 46. control in children with brain tumors. Neurosurgery 37 Wassermann EM, Pascual-Leone A, Hallett M. Cortical 1989;25:786–92. representation of the ipsilateral hand and arm. Exp Brain 7 Ebeling U, Schmid UD, Ying H, et al. Safe surgery of lesions Res 1994;100:121–32. near the motor cortex using intra-operative mapping 38 Kim SG, Alshe J, Georgopoulos AP, et al. Functional imag- techniques: a report on 50 patients. Acta Neurochir (Wien) ing of the human motor cortex at high magnetic field. J 1992;119:23–8. Neurophysiol 1993;69:297–302. 8 Seitz RJ, Huang Y, Knorr U, et al. Large-scale plasticity of 39 Shibasaki H, Sadato N, Lyshkow H, et al. Both primary the human motor cortex. Neuroreport 1996;6:742–4. motor cortex and supplementary motor area play an 9 Atlas SW, Howard II RS, Maldijian J, et al. Functional mag- important role in complex finger movement. Brain netic resonance imaging of regional brain activity in 1993;116:1387–98. patients with intracerebral gliomas: findings and implica- 40 Weiller C, Ramsay SC, Wise RJ, et al. Individual patterns of tions for clinical management. Neurosurgery 1996;38:329– functional reorganization in the human cerebral cortex 38. after capsular infarction. Ann Neurol 1993;33:181–9. 10 Roux FE, Boulanouar K, Ranjeva JP, et al. Usefulness of 41 Werring DJ, Clark CA, Barker GJ, et al. The structural and motor functional MRI correlated to cortical mapping in functional mechanisms of motor recovery: complementary rolandic low-grade astrocytoma. Acta neurochir (Wien) useofdiVusion tensor and functional magnetic resonance 1999;140:71–9. imaging in a traumatic injury of the internal capsule. J 11 Fandino J, Kollias SS, Wieser HG, et al. Intraoperative vali- Neurol Neurosurg Psychiatry 1998;65:863–9. dation of functional magnetic resonance imaging and cor- 42 Schlosser MJ, McCarthy G, Fulbright RK, et al. Cerebral tical reorgaization patterns in patients with brain tumors vascular malformations adjacent to sensorimotor and involving the primary motor cortex. J Neurosurg 1999;91: visual cortex. Functional MRI studies before and after 238–50. therapeutic intervention. Stroke 1997;28:1130–7. 12 Latchaw RE, Hu X. Functional MR imaging in the evalua- 43 Cao Y, Vikingstad EM, Huttenlocher PR, et al. Functional tion of the patient with epilepsy. Functional localization. magnetic resonance studies of the reorganization of the Neuroimaging Clin N Am 1995;5:683–93. human hand sensorimotor area after unilateral brain injury 13 Stapleton SR, Kiriakopoulos E, Mikulis D, et al. Combined in the perinatal period. Proc Natl Acad Sci USA utility of functional MRI, cortical mapping and frameless 1994;91:9612–16. stereotaxy in the resection of lesions in eloquent areas of 44 Maegaki Y, Yamamoto T, Takeshita K. Plasticity of central http://jnnp.bmj.com/ brain in children. Pediatr Neurosurg 1997;26:68–82. motor and sensory pathways in a case of unilateral 14 Yoshiura T, hasuo K, Mihara F, et al. Increased activity of extensive cortical dysplasia. Neurology 1995;45:2255–60. the ipsilateral motor cortex during a hand motor task in 45 Cramer SC, Nelles G, Benson RR, et al. A functional MRI patients with brain tumors and paresis. Am J Neuroradiol study of subjects recovered from hemiparetic stroke. Stroke 1997;18:865–9. 1997;28:2518–27. 15 Caramia MD, Telera S, Palmieri MG, et al. Ipsilateral motor 46 Wunderlich G, Knorr U, Herzog H, et al. Precentral glioma activation in patients with cerebral gliomas. Neurology location determines the displacement of cortical hand rep- 1998;51:196–202. resentation. Neurosurgery 1998;42:18–27. 16 Côté R, Hachinski VC, Shurvell BL, et al. The Canadian 47 Kim SG, Alshe J, Hendrich K, et al. Functional magnetic neurological scale: a preliminary study in acute stroke. resonance imaging of motor cortex: hemispheric asymme- Stroke 1986;17:731–7. try and handedness. Science 1993;261:615–17. 17 Bastings EP, Rapisarda G, Pennisi G, et al. Mechanisms of 48 Li A, Yetkin Z, Cox R, et al. Ipsilateral hemisphere activation on October 1, 2021 by guest. Protected copyright. hand motor recovery after strokes: an electrophysiologic during motor and sensory tasks. AJNR Am J Neuroradiol study of central motor pathways. J Neurorehab 1997;11:97– 1996;17:651–5. 108. 49 Jones RD, Donalson IM, Parkin PJ. Impairement and recov- 18 Friston KJ, Holmes AP, Worsley KJ, et al. Statistical ery of ipsilateral sensory-motor function following unilat- parametric maps in functional imaging: a general appoach. eral cerebral infarction. Brain 1989;112:113–32. Hum Brain Mapp 1995;2:189–210. 50 Netz J, Lammers T, Homberg V. Reorganization of motor 19 Talairach J, Turnoux P. Co-planar stereotactic atlas of the output in the non-aVected hemisphere after stroke. Brain human brain. New York: Thieme Medical, 1988. 1997;120:1579–86. 20 Puce A, Constable RT, Luby ML, et al. Functional magnetic 51 Zarzecki P, Witte S, Smits E, et al. Synaptic mechanisms of resonance imaging of sensory and motor cortex: compari- cortical representational plasticity: somatosensory and cor- son with electrophysiological localization. J Neurosurg ticocortical EPSPs in reorganized raccoon SI cortex. J 1995;83:262–70. Neurophysiol 1993;69:1422–32. 21 Rao SM, Bandettini PA, Binder JR, et al. Relationship 52 Jacobs KM, Donoghue JP. Reshaping the cortical motor between finger movement rate and functional magnetic map by unmasking latent intracortical connections. Science resonance signal change in human primary motor cortex. J 1991;251:944–7. Cereb Blood Flow Metab 1996;16:1250–4. 53 Cohen LG, Roth BJ, Wasserman EM, et al. Magnetic stimu- 22 Segebarth C, Belle V, Delon C. Functional MRI of the lation of the human cerebral cortex, an indicator of human brain: predominance of signals from extracerebral reorganization in motor pathways in certain pathological veins. Neuroreport 1994;5:813–16. conditions. J Clin Neurophysiol 1991;8:56–65. 23 Sergent J. Brain-imaging studies of cognitive functions. 54 Levy WJ, Amassian VE, Schimd UD, et al. Mapping of TINS 1994;17:221–7. motor cortex gyral sites non-invasively by transcranial 24 Türck L. Uber secundäre erkrankung einzlner rückenmark- magnetic stimulation in normal subjects and patients. In: stränge und ihrer fortsetzung zum gehirne. SB Akad Wiss Levy WJ, Cracco RQ, Barker AT, et al,eds.Magnetic motor Wein 1851;6:288. stimulation: basic principles and clinical experience (EEG 25 Holmes GL, May WP. On the exact origin of the pyramidal suppl). Amsterdam: Elsevier, 1991. tracts in man and other mammals. Brain 1909;32:1–42. 55 Sanes J, Donoghue J. Organization and adaptability of mus- 26 Kuypers HG. Anatomy of the descending pathways. In: VB cle representation in the primary motor cortex. Exp Brain Brooks, ed. Handbook of physiology, section 1: the nervous sys- Res 1992;22(suppl):103–7.

www.jnnp.com The Fissure of Sylvius (1614–72) 463 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.69.4.453 on 1 October 2000. Downloaded from

56 Huttenlocher PR, Raichelson RM. EVects of neonatal 58 Dettmers C, Fink GR, Lemon RN, et al. Relation between hemispherectomy on location and number of corticospianl cerebral activity and force in the motor areas of the human neurons in the rat. Dev Brain Res 1989;47:59–69. brain. J Neurophysiol 1995;74:802–15. 59 Turton A, Wroe S, Trepte N, et al. Controlateral and ipsilat- 57 Marque Ph, Felez A, Puel M, et al. Impairement and recov- eral EMG responses during recovery of arm and hand ery of left motor function in patients with right hemiplegia. function after stroke. Electroencephalogr Clin Neurophysiol J Neurol Neurosurg Psychiatry 1997;62:77–81. 1996;101:316–28.

Medicine. He gave his inaugural lecture De His works in neuroanatomy were published in Hominis Cognitione on 17 September. He was the disputation De spirituum animalium in cer- HISTORICAL NOTE an excellent and enthusiastic teacher and lec- ebro, cerebelloque confectione, per nervos distribu- turer, concentrating on the more common tione, atque usu vario, defended by the student ailments prevalent at the time in the Caecilia Gabriel Ypelaer under Sylvius’s supervision Hospital. He applied the Socratic method in 1660. The lateral fissure of Sylvius is and emphasised the modern systems of diag- described: The fissure of Sylvius (1614–72) nosis, prognosis, and therapy. His work shows ” . . .the surface of the cerebrum is very the importance he attached to necropsies as a deeply marked by twistings (gyri) which Franciscus called Sylvius, was a descendant way of verifying or rejecting clinical diag- are somewhat similar to convolutions of of a protestant family named Du Bois noses, as well as giving clues as to the nature the small intestine. Particularly noticeable (changed to de le Boë1) from Cambrai, in of the disease. It is Sylvius who is said to have is the deep fissure or hiatus which . . .be- France. For religious reasons the family first demonstrated at necropsy the lung gins at the roots of the eyes(oculorum moved to Germany. Sylvius was born in tubercles. radices) . . .it runs posteriorly above the Hanau, Germany. He read medicine at the Sylvius became interested in iatrochemis- temples as far as the origin of the brain stem (medullae radices) . . .It divides the universities of Sedan and Leiden. He began try, a concept that sought to explain physi- his studies in June, 1632 at Leiden and cerebrum into an upper, larger part and a ological processes as dependent on chemical lower, smaller part. Twistings occur along oVered a disputation Positiones variae medicae mechanisms; the idea was not far removed in 1634. Sylvius obtained his medical doctor- the fissure’s length and depth even with from modern neurotransmitters. Of the the origins of smaller convolutions at the ate at the University of Basel on 16 March theses presented under his presidency, one most superior part of it.” (1663, pp 1637, defending a thesis De animali motu named Disputationem medicarum de- 43–44.) ejusque laesionibus. According to Haller this is cas1(1663), contained “the primary natural Sylvius’s accurate study of the outer surface the first description of the lateral cerebral fis- functions of the human body deduced from of the brain was done because of his interest sure. This fissure and the cerebral aqueduct anatomical, practical and chemical experi- in the vascular system on the brain surface, were not fully described by Sylvius until ments.” Sylvius regarded as fundamental the 1 and his interest in the grey matter related to 1663. eVervescence or violent reaction, between the animal spirits. He practised for a short period but gradu- acid and alkaline secretions and he rejected ated again at Leiden University in November the classic humours. But he retained the The ventricular aqueduct 1638. His skills in teaching anatomy brought notion of the animal spirits. These spirits in The connection between the third and fourth him respect and a certain fame: “many the blood were transported by the neck arter- ventricles had already been mentioned or students, and certainly not the worst ones, ies in the capillaries on the brain surface in a supposed by Galen in as a attended his courses, so that it seemed as if process analogous to distillation. The most De usum partium canal giving communication between the cer- only he could understand and explain spiritual part of the blood passed the pores of ebrum and the cerebellum. Vesalius had anatomy.” capillaries, first in the grey matter and then in described in the (1543) an “anus-like One of these students was Thoma Bartho- the white matter. Fabrica orifice of the meatus which extends from the lini, son of the famous Danish anatomist Because of his brilliance, clinical teaching third to the fourth ventricle” below the quad- Caspar Bartholini. In the 1641 edition of his in Leiden flourished under Sylvius and rigeminal bodies. In chapter 21 in the dispu- well known textbook Institutiones anatomicae attracted many students from many other

tation of Franciscus Sylvius is described a http://jnnp.bmj.com/ published by Thoma, 12 years after Caspar’s countries. When he died on 15 November canalis vel aquae-ductus between the con- death, it is clear that Caspar with Sylvius had 1672, the medical faculty of Leiden went into joined roots of the spinal cord and under “our shown and named the cerebral fissure sepa- a relative state of decline. bridge” (pons Varoli) and the corpora rating the temporal lobe from the frontal lobe Sylvius published his pathology under the 2 quadrigemina. The aqueduct was certainly above. Not until 1663 was it separately pub- title . Unfortunately, Praxeos medica idea nova known before Sylvius, and both Haller and lished in Sylvius’s Opera as Disputationes he could only complete the first volume Morgagni were critical of naming the aque- medicarum ad C Bartholini Institutiones Ana- (1671). His former pupil Justus Schrader duct after Franciscus Sylvius. tomicas, but Caspar Bartholini gave credit to posthumously published the other volumes, The naming of the fissure followed Thomas Sylvius for the discovery, probably in his the- including the appendix. Generally speaking, Bartholini’s homage to Sylvius’s work, but it sis of 1637. diseases were caused by abnormal eVerves- on October 1, 2021 by guest. Protected copyright. had been described by Caspar Bartholini. After this important work, Sylvius turned cence due to abnormal secretions, which his attention to the circulation. He was able to could be either sharp alkaline or sharp acidic. J M S PEARCE show that the blood had an independent flow A defective animal spirit resulting from an 304 Beverley Road, Anlaby, or circulation through the blood vessels, accumulation of a volatile acid spirit, for Hull HU10 7BG, UK pumped by the heart. Thus, he tried to con- instance, caused epilepsy. Therapy consisted Correspondence to: [email protected] vince his seniors of the medical faculty that of the prescription of alkaline salts, opposing Harvey’s theory was correct. Professor Johan the action of the excess of acid. Walaeus, one of his professors, became a His contributions to the anatomy of the 1 Sylvius F. (de la Boë) Disputationes medicarum spirited proponent of Harvey’s theory. Syl- pars prima. Amstelodami, van den Bergh, brain were recognised by Thomas Bartholi- 1663. Cited by: McHenry LC. Garrison’s vius had proposed the circulation in the lungs nus, who in 1640 remarked: History of neurology. Springfield, Illinois: in his thesis of 1634, 6 years after De motu ”we can not pass over in silence the very Charles C Thomas 1969:64. cordis. accurate anatomist D. Franciscus Sylvius, 2 Bartholin C. Institutiones anatomicae, novis recen- Sylvius moved from Leiden to Amsterdam tiorum opinionibus and observationibus quarum since we borrow from his noble brain and innumerae hactenus editae non sunt, figurisque in 1641. He practised there for 17 years, until ingenuity the admirable new structure of auctae ab auctoris filio Thoma Bartholino. Lug in 1658 he returned to Leiden as Professor of the brain.” Batavorum, Apud Franciscum Hackium, 1641.