British Journal of Neurosurgery, 2015; Early Online: 1–9 © 2015 The Neurosurgical Foundation ISSN: 0268-8697 print / ISSN 1360-046X online DOI: 10.3109/02688697.2015.1006170

ORIGINAL ARTICLE

Obtaining the olfactory bulb as a source of olfactory ensheathing cells with the use of minimally invasive neuroendoscopy-assisted supraorbital keyhole approach—cadaveric feasibility study Marcin Czyz1,2*, Pawel Tabakow1*, Irene Hernandez-Sanchez3, Wlodzimierz Jarmundowicz1, & Geoffrey Raisman4

Br J Neurosurg Downloaded from informahealthcare.com by Nyu Medical Center on 04/24/15 For personal use only.

1Department of Neurosurgery, Wroclaw Medical University, Wroclaw, Poland, 2The Centre for Spinal studies and Surgery,

Nottingham University Hospitals, UK, 3Faculty of Medicine, Universitat Autònoma de Barcelona, Barcelona, Spain, and 4Spinal Repair Unit, UCL Institute of Neurology, London, UK

due to the limited neurogenesis and the lack of spontaneous regrowth ability of lesioned central axons. Factors determining the observed abortive regeneration in the CNS are the inhibitory influence of central posttraumatic glial scar and myelin-associated proteins such as Nogo, the lack of intrinsic capacity of surviving neurons for regeneration, and the lack of an effective glial pathway to guide regenerating axons across the CNS lesion.1 Olfactory ensheathing cells (OECs) are a unique population of glial cells found in the lamina propria of the olfactory mucosa, in the bundles of olfactory filia, and in the two outer layers of the olfactory bulb (OB).1,2 OECs have been shown to evoke anatomical and functional regeneration in various animal models of spinal cord and brachial plexus injury.3–5 According to the results of laboratory and clinical observations, OECs promote the restoration of neurological function of the injured spinal cord increasing the plasticity of neural tissue and stimulating the regeneration of long spinal cord tracts.6–8 The OECs can be isolated either from the olfactory mucosa or the OB. While in majority of preclinical experiments focused on the repair of the injured mammalian spinal cord, OECs were obtained from the OB,5,9–11 the olfactory mucosa is the most frequently used source of these cells in clinical trials mainly due to the relatively well accessibility and the ability of the olfactory tissue for self-renewal.8,12–15 Yet recent studies have shown that the OB-derived OECs have higher potential to promote neuroregeneration when compared with olfactory mucosa.3,16–18 This finding raises the question if the olfactory mucosa could be safely replaced by the OB as a source of OECs for therapeutic use. The obtaining of biopsies from the olfactory mucosa is a minimally invasive procedure. The nasal cavity is easily accessible; the postoperative restoration of the olfaction is spontaneous and fast; and operations are safe, reproducible, and well tolerated by patients.8,13,14 The main drawbacks

Abstract Background. Obtaining the human olfactory bulb (OB) for treatment of spinal cord injuries with olfactory ensheathing cells (OECs) requires the elaboration of a surgical approach that could meet the criteria of safety and minimal invasiveness. The aim of the study was to evaluate the suitability of the keyhole supraorbital craniotomy (SOC) with an eyebrow incision for obtaining OB for therapeutic purposes. Methods. Seventeen SOCs were performed on nine fresh adult cadavers. The procedure of obtaining OB was conducted by neuroendoscopeassisted microsurgical dissection. Technical features related to the procedure were measured and adverse events were noted. The virtual three-dimensional planning was applied in six cases to verify the authorial A–C scale published previously. Results. The intact OB was obtained in 10 (59%) cases and a minor injury was discovered in another 5 (29%) cases. In 2 (12%) specimens OB was severely damaged which was correlated with the minor neural tissue injury (Fi2 ⫽ 0.44). While no case of an evident vascular injury was noted, there were 3 (18%) incidents of unintended frontal sinus opening positively correlated with the craniotomy width (Fi2 ⫽ 0.44). The unfavorable threedimensional (3D) configuration of the olfactory groove area was revealed in 66% of cases and highly correlated with OB injury (Fi2 ⫽ 1.0) but not damage. Conclusions. The SOC via an eyebrow incision may be safely and effectively applied to obtain the OB as a source of OECs in roughly 90% of cases. Virtual 3D planning is useful in preoperative qualification of potential donors. Keywords: minimally invasive neurosurgery; neuroregeneration; olfactory bulb; olfactory ensheathing cells; spinal cord injury

Introduction Injuries of the central nervous system (CNS) in humans often lead to irreversible neurological impairment and disability

*Both the authors contributed equally to this work. Correspondence: Marcin Czyz, MD, PhD, Department of Neurosurgery, Wroclaw Medical University, Ul. Borowska 213, 50-556 Wrocław, Poland. Tel: ⫹ 48 071 734-34-00. Fax: ⫹ 48 071 734-34-90. E-mail: [email protected]. Received for publication 8 August 2014; accepted 4 January 2015

1

Br J Neurosurg Downloaded from informahealthcare.com by Nyu Medical Center on 04/24/15 For personal use only.

2

M. Czyz et al.

attributed to the retrieval of olfactory tissue are the potential risk of bacterial contamination, small number of available OECs, and the presence of stem cells and numerous other supportive mucosal cells among the desired OECs, which may cause uncontrolled cell proliferation.19 In contrast, the OB predominantly contains cells from neural origin, including numerous of OECs and is situated in the sterile intradural space. The main and obvious disadvantage of obtaining bulbar OECs is the necessity to perform an operative approach to the anterior skull base and a unilateral bulbectomy, which raises legitimate concerns about the invasiveness and the potential risks of the procedure.20–22 The obtaining of the human OB for treatment of spinal cord injuries with OEC transplants requires the elaboration of a surgical approach that could meet the criteria of safety and minimal invasiveness, and would be acceptable by the patients.23,24 Previously we proposed the keyhole supraorbital craniotomy (SOC) with an eyebrow incision to obtain the OB for therapeutic purposes.25 The three-dimensional (3D) virtual modeling allowed us to state that it might be a promising method of harvesting bulbar OECs of good quality with a well-known minimal risk of perioperative morbidity and an excellent esthetical outcome.26–28 Although the results obtained were encouraging, there is a need to perform a cadaveric feasibility study prior to introducing the method to a routine clinical practice.29,30 The purpose of this study was to verify the suitability of the SOC basing on the cadaver simulation of the procedure accomplished with the assessment of the quality of OBs obtained. An additional validation of the preoperative 3D planning method presented previously was also performed.

Materials and methods Cadaveric study Between February and August 2013 simulated operations of the OB harvesting were performed in the Cadaver Dissection Laboratory developed on the basis of the local department of pathology. The laboratory has been equipped with microneurosurgical instruments including toolset dedicated

for transcranial neuroendoscope-assisted microneurosurgery (Aesculap AG, Germany), Midas Rex high-speed drill (Medtronic, Inc., USA), suction device (Skimed, Poland), Mayfield Infinity skull clamp (Integra, USA) accomplished with custom-made mounting device, NC 31 operative microscope (Carl Zeiss, Germany), and neuroendoscope with a 0° and 30° angle of view (Aesculap AG, Germany) (Fig. 1). The exclusion criteria were documented paranasal sinuses, orbit or anterior skull base pathologies as well as facial deformations. Fresh cadavers were stored in the refrigerator at temperature of 3.3°C (38°F) for less than 48 h and simulated operations were performed prior to the autopsy. The study was approved by the Bioethics Committee of Wroclaw Medical University, according to the guidelines of the National Health Council of Poland in adherence to the Declaration of Helsinki. According to the national regulations every postmortem performed at a teaching institution may be accomplished with didactic or scientific procedures as long as those do not impair the esthetical integrity and are not related to irretrievable obtaining specimens for external examinations (e.g., histology and anatomical specimens or casts). Nevertheless, in every case prior to the autopsy patient’s next-of-kin is informed about the scientific nature of the postmortem and formal consent is obtained. Our study was performed at the University Hospital and designed in order to meet all of the above-mentioned criteria hence accepted by the Local Bioethics Committee. The keyhole SOC with an eyebrow skin incision was performed on each side according to guidelines introduced by Perneczky and Reisch.28,31,32 As the cadaver laid down in a supine position the head was placed in the Mayfield headholder, elevated, retroflected by about 15° and rotated by ca. 60°. After head positioning the operating field was draped and the orbital rim, supraorbital foramen or notch, temporal line, and the zygomatic arch were identified by palpation (Fig. 2A). The skin incision started laterally from the supraorbital incision until the temporal line and the skin flap were dissected in the frontal direction. The frontal muscle was cut ca. 2 cm over the orbital rim. The visible part of the temporalis muscle was stripped from the bony insertion and the musculo-cutaneus

Fig. 1. Cadaver laboratory. (A) Mayfiled’s headholder with a custom-made mounting system based on technical window glass holder suckers (white star). (B) Headholder attached to the autopsy table. Notice that all degrees of freedom of the Mayfield’s headholder are available. (C) Full setting of the dissecting room—cadaver placed on the autopsy table and fully draped, endoscopy toolset (white star) and operative microscope (black star) are in standard positions.

Br J Neurosurg Downloaded from informahealthcare.com by Nyu Medical Center on 04/24/15 For personal use only.

Obtaining OEC via supraorbital craniotomy 3 flap was finally retracted with holding sutures in frontal and lateral direction (Fig. 2B). Single fronto-basal burr hole was performed laterally from the temporal line at the level of the fronto-zygomatic suture. The first craniotomy cut was parallel to the orbital rim in a lateral to medial direction, the second one—“C” shaped—extending from the burr hole to the median border of the first line (Fig. 2A). Once bone flap was removed, the width and height of the craniotomy were measured and the operating microscope was introduced. The internal edge of the bone above the orbital rim as well as juga cerebralia were removed by highspeed drilling increasing the angle for visualization and manipulation. The dura was opened in a “C”-shaped manner and retracted in basal direction (Fig. 2C). The frontal lobe was protected with a patty and carefully elevated with a retractor. After identification of the optic nerve and carotid artery relevant arachnoid cisterns were opened and cerebrospinal fluid (CSF) was sucked out increasing the mobility of neural structures. The retractor was removed and the olfactory tract (OT) identified, dissected from the basal surface of the frontal lobe, and cut in its proximal part.

In the next step, the rigid neuroendoscope with 4-mm diameter and 30° viewing angle was introduced into the operating field. The optic nerve, internal carotid artery, and cut OT were identified (Fig. 3A). The OT was continuously mobilized out of the frontal lobe beginning from the cutting site, down to the OB (Fig. 3B). The bony eminence limiting lateral aspect of the OG was drilled out if needed. In all cases an attempt to obtain an intact OB was made by combining a blunt and sharp dissection (Fig. 3C). The olfactory filia were cut as distally to the OB as possible. The distal part of the OT was subsequently localized and grasped. The OB–OT complex was then carefully removed from the operating field and stored in 0.9% NaCl solution at the room temperature until the end of the procedure. The OG area was inspected for the possible OB remnants and the operating field was checked for any visible injuries of vascular or neural structures (Fig. 3D). To assess the length of the instruments, transition trajectory of the distance between upper edge of the craniotomy and the middle part of OG was measured with a thin rigid probe. The dura was closed with the interrupted sutures and osteoplastic wound closure was performed.

Fig. 2. (A) Image of the right-sided keyhole SOC (dotted line) with eyebrow skin incision marked on a male cadaver fused with a 3D skull model using GNU image manipulation program: GIMP 2.8.10. Note burr hole performed laterally to superior temporal line (asterisk) at the level of frontozygomatic suture (black arrow). The medial extent of the craniotomy is limited by the supraorbital notch (white arrow). (B) Skin flap was dissected in the frontal direction and the frontal muscle was cut ca. 2 cm over the orbital rim. The visible part of the temporalis muscle was stripped from the bony insertion and the musculo-cutaneus flap was finally retracted with holding sutures in frontal and lateral direction. Note temporal line and temporal fossa visible in right aspect of the bony exposure. (C) The craniotomy was introduced after single fronto-basal burr hole and blunt dura mater mobilization. The internal edge of the bone above the orbital rim as well as juga cerebralia were removed by high-speed drilling, increasing the angle for visualization and manipulation. The dura was opened in a “C”-shaped manner and retracted in basal direction.

Br J Neurosurg Downloaded from informahealthcare.com by Nyu Medical Center on 04/24/15 For personal use only.

4

M. Czyz et al.

Fig. 3. The 30° endoscopic view during OB harvesting via a right-sided keyhole SOC. (A) Proximal (black arrow) and distal (white arrow) parts of the cut OT on the foreground. The optic nerve (black asterisk) and internal carotid artery (white asterisk) are also exposed. (B) The OT (black arrow) elevated with the dissector, the OB (white arrow) is visible before being dissected from the OG (white asterisk). (C) Higher magnification of a partially dissected OB and OB removed from the OG. (D) The inspection of the OG—internal surface of the cribriform plate with ethmoid foramina (black arrow) and OB remnants are visible.

The intracutaneous running suture was routinely applied. The “skin-to-skin” time of the operation (Top) was measured and noted. The operator’s comfort was assessed in 10-point subjective scale—equivalent to Visual Analog Scale for pain assessment—where 1 meant no comfort at all and 10 meant full comfort.33 All incidents of the frontal sinus opening, OB, neural and vascular structures injury, as well as a need for an additional skull base drilling in the OG area were noted. The quality of the OB obtained was assessed just after the operation under the operative microscope by the author experienced in obtaining of viable OBs for OEC isolation and culture (P. T.).18 Three classes of the OB obtained were distinguished—(1) Intact OB. No visible damage, one fragment; (2) OB injury. Minor damage visible, no more than two fragments; and (3) OB damage. Severe damage or greater tissue fragmentation (Fig. 4). Both the first authors, experienced in classical microneurosurgical and endoscopy-assisted procedures via the SOC, performed all operations.

Computer simulation In three patients, premortem CT examinations—which were performed following the same 3D volume axial helical

protocol with a 0.625-mm slice thickness using a spiral Dual HiSpeed (GE Healthcare, United Kingdom) CT unit with a gantry tilt of zero degrees—were available. In these cases, a 3D computer simulation of the OB harvesting were performed following the objective and reproducible protocol described in details previously.25 Briefly, the 3D model of the skull was prepared for each individual case in the neuronavigation workstation (Cranial 5, StealthStation Application Software, Medtronic Navigation). The virtual keyhole SOC was placed on each side of the model with respect to the classic anatomical landmarks. Trajectories of neurosurgical instrument transitions to the anterior and posterior aspects of olfactory grooves (OG) were subsequently designed and measured with correction allowing the avoidance of collisions with bony eminences of the skull base. A standard option trajectory was used, which is offered by software and used in daily neurosurgical practice. After mapping the craniotomy, the entry point was located in the middle of its horizontal aspect, 20 mm above the base. The anterior and posterior target points were marked at the bottom of the OG. The connection of the entry point with the target points created two lines, which indicated the trajectory of the neurosurgical instrument transition. Subsequently, the

Fig. 4. Three classes of the obtained OB were distinguished—(A) Intact OB. No visible damages, one fragment. (B) OB injury. Minor damages visible, no more than two fragments. (C) OB damage. Severe damages are visible.

Br J Neurosurg Downloaded from informahealthcare.com by Nyu Medical Center on 04/24/15 For personal use only.

Obtaining OEC via supraorbital craniotomy 5 trajectories were analyzed with the use of a Probe’s eye view option of the neuronavigation software, which provides a plane of view, which is perpendicular to the trajectory, and allows visualizing potential conflict of the imaginary tool with critical structures (e.g., vessels) during preoperative planning of stereotactic procedures. In our case, we study the trajectories for possible collisions with skull base structures, such as protuberances, digital impressions of the anterior fossa, and the edges of the OG. In case of the occurrence of such obstacles, another collision-free trajectory was calculated by mapping a new target point. This was done by raising the destination points in the vertical axis to the closest possible spot that provided a clean transition just above the cranial structures. Trajectory lengths and the distances between previous destination points and new collision-free target points—called “corrections”—were measured. Each side was treated as a separate case and classified into one of nine types of OG anatomical configuration, reflecting the accessibility of the OB in the OG area. Each case was marked by two letters (e.g., CA). The anterior and the posterior aspects of the OG were assigned the first and second letters, respectively. Type A—OB localized superficially, there was no need to correct the initial trajectory to the base of OG; type B—OB partially hidden, there was a need to correct the initial trajectory, but the correction value did not exceed the depth of the OG; type C—OB totally hidden, unfavorable configuration from the surgical point of view, the correction value was greater than the depth of the OG. A simulation of the approach, measurements, and the assignment was performed in each case by an investigator blinded to the results of the cadaveric simulation.

Statistical analysis The gathered data were compared and analyzed statistically with Statistica 10 (StatSoft, Inc.) and MedCalc 12 (MedCalc Software bvba); statistical significance was defined as a type-I error ⬍0.05. After testing for normal distribution

using the Shapiro–Wilk W test, the Mann–Whitney U test was applied to check possible asymmetry as well as operatorand gender-related differences. Correlations between specific measurements and features were tested with Fi2 test. An analysis of change in Top, and events of the OB and neural structures damage in relation to number of cases operated was performed on the basis of dot diagrams complemented by multinomial and linear trends. A multivariate analysis of factors listed in Table I—potentially related to injury or damage of the OB—was performed with the use of a logistic regression models. Only factors with the p ⬍ 0.2 in the univariate analyses were selected for analysis in the final model.

Results Cadaveric study Between February and August 2013, seventeen (nine on the right and eight on the left) simulated operations of the OB harvesting were performed on nine fresh adult non-preserved cadavers obtained from routine autopsies (Table I). In one case (CD05) it was not possible to perform left-sided craniotomy due to the organizational reasons (time limitation). The mean age of the study group was 72 ⫾ 5 years, 4 (44%) subjects were male. None of the specimens presented with any signs of an extensive brain edema, autolysis, or any major pathology of the anterior skull base and frontal lobes. There were no significant differences related to either the side of the operation or gender. Mean craniotomy width and height were 29.53 ⫾ 3.32 mm and 19.29 ⫾ 2.23 mm, respectively. The mean Top was 78 ⫾ 22 min and the mean trajectory length was 54.71 ⫾ 5.68 mm. The median operator’s comfort noted was 8 and varied between 3 and 10. The intact (class 1) OB was obtained in 10 (59%) cases and a minor injury (class 2) was discovered in another 5 (29%); in 2 (12%) cases OB were severely damaged (class 3). No case of an evident vascular injury was noted. A minor injury of basal part of the

Table I. Simulated operations of the OB harvesting performed via eyebrow incision minimally invasive SOC. N ⫽ 17

Sex

Age

Side

W

H

Top

Traj

Neur

Vesel

OBI

OBD

Sinus

Drill

Comf

CD01

M

72

CD02

F

67

CD03

M

84

CD04

F

66

L R L R L R L R

30 30 30 30 25 27 28 33

20 20 20 17 17 20 17 18

120 100 120 105 60 90 60 70

53 54 55 53 57 54 50 54

1 1 0 0 0 0 0 1

0 0 0 0 0 0 0 0

1 0 1 1 0 0 0 0

0 0 0 0 0 0 0 0

0 0 1 1 0 0 0 1

0 1 1 1 1 0 0 0

6 6 8 8 8 9 10 3

CD05

F

73

CD06

M

71

CD07

F

74

CD08

M

67

CD09

F

73

R L R L R L R L R

44%M

72 ⫾ 5

23 29 30 30 29 37 35 28 28 29.53 ⫾ 3.32

17 23 17 18 20 25 20 19 20 19.29 ⫾ 2.23

90 70 60 55 65 60 80 61 60 78 ⫾ 22

60 57 49 60 45 58 68 58 45 54.71 ⫾ 5.68

0 1 0 1 1 0 1 0 0 41%

0 0 1 0 0 0 0 0 1 29%

0 1 0 0 0 0 1 0 0 12%

0 0 0 0 0 0 0 0 0 18%

0 0 0 1 1 0 1 0 1 47%

9 7 7 9 8 8 6 8 6 8 (3–10)

Summary

0 0 0 0 0 0 0 0 0 0%

3D

CA CA

BA CA

BA CA

M, male; F, female; W, craniotomy width [mm]; H, craniotomy height [mm]; Top, time of the operation [min]; Traj, trajectory length [mm]; Neur, neural structures’ damage; Vessel, vascular structures’ damage; OBI, olfactory bulb injury; OBD, olfactory bulb damage; Sinus, unintended frontal sinus opening; Drill, excessive skull base drilling needed; Comf, operator’s comfort (1–10); 3D, head CT available and 3D planning results.

6

M. Czyz et al. Table II. Correlation matrix for parameters assessed. Only statistically significant (p ⬍ 0.05) Fi2 coefficient values are presented. H˅

Br J Neurosurg Downloaded from informahealthcare.com by Nyu Medical Center on 04/24/15 For personal use only.

H˅ W˅ H˄ W˄ T˅ T˄ Neur Drilling Sinus Comf˅ Traj˄ OBI OBD

W˅

H˄

W˄

T˅

T˄

Neur

Drill

Sinus Comf˅ Traj˄

OBI

OBD

– ⫺0.62

– – –

0.44 – –

⫺0.62

–

0.51

0.44

– 0.44

– 0.51

–

0.44 – –

0.44

0.44

–

H˅, craniotomy height lower than average; W˅, craniotomy width lower than average; H˄, craniotomy height exceeding the average; W˄, craniotomy width exceeding the average; T˅, time of the operation shorter than average; T˄, time of the operation exceeding the average; Neur, neural structures damage; Drill, excessive skull base drilling needed; Sinus, unintended opening of the frontal sinus; Comf˅, operator’s comfort lower than average; Traj˄, the distance between upper edge of the craniotomy and the middle part of the olfactory groove exceeding the average; OBI, olfactory bulb injury (class 2); OBD, olfactory bulb damage (class 3).

frontal lobe (arachnoid tear) was observed in 7 (41%) cases which was less common for craniotomies when the width was lower than average (P ⫽ 0.003), and was related to lower than average operator’s comfort (P ⫽ 0.031) and OB damage (P ⫽ 0.046) but not injury (Table II). There were 3 (18%) incidents of unintended frontal sinus opening positively correlated with the craniotomy width (P ⫽ 0.036). The Top gradually decreased from 120 to reach the plateau of roughly 70 min after eight out of seventeen operations were performed. There were no noted correlations between OB or neural structures damage and number of cases operated (Fig. 5).

Computer simulation An unfavorable 3D configuration of the OG area was revealed during preoperative analysis in 4 (67%) of 6 cases. In all of these cases, a minor (class 2) OB injury or no incidence of OB damage were noted. An unfavorable 3D configuration was related to the shorter trajectory, the need of an extensive skull base drilling, and class-2 OB injury (Table III).

Logistic regression model For the OB damage on univariate analysis, unfavorable prognostic factors were (P ⬍ 0.1) operator’s comfort below average, craniotomy height larger than average, an iatrogenic

Fig. 5. Upper row: learning curve reaching the 70-min plateau after eight out of seventeen operations were performed. Lower row: no essential correlation between OB (on left) or neural structures (on right) damage and number of cases operated.

Obtaining OEC via supraorbital craniotomy 7 Table III. The Fi2 coefficient and p values noted for the relation of unfavorable 3D configuration of the OG area revealed by virtual planning with particular features. Detailed description of the method used is in the text. [*].

3D˅

Top

Neur

Drill

Sinus

Comf˅

Traj

OBI

OBD

P ⫽ 0.14

P ⫽ 0.10

0.70 P ⴝ 0.05

P ⫽ 0.15

P ⫽ 1.0

ⴚ1.0 P ⴝ 0.02

1.0 P ⴝ 0.01

P ⫽ 0.12

3D˅, unfavorable 3D configuration of the skull base in the OG area; Top, time of the operation; Neur, neural structures damage; Drill, excessive skull base drilling needed; Sinus, unintended opening of the frontal sinus; Comf˅, operator’s comfort lower than average; Traj, the distance between upper edge of the craniotomy and the middle part of the olfactory groove; OBI, olfactory bulb injury (class 2); OBD, olfactory bulb damage (class 3). Statistically significant results were put in bold font.

neural tissue injury, and the trajectory longer than average. The multivariate analysis revealed the significant overall model fit (p ⫽ 0.025).

Br J Neurosurg Downloaded from informahealthcare.com by Nyu Medical Center on 04/24/15 For personal use only.

Discussion Results from a large number of preclinical and clinical experiments focused on the functional regeneration of lesioned spinal cord axons after transplantation of OECs showed that the OB seems to be the best source of these.16,34,35 While there is extensive information from the literature about the successful application of OECs isolated from the OB of different mammals (rats, dogs, cats, and monkeys), there are very few reports concerning the retrieval of the human OB for the isolation and culture and transplantation of OECs.18,34–36 In our previous paper we presented a radiological feasibility study based on a virtual 3D modeling with the aim of simulating the minimally invasive approach to the OG with the use of the keyhole SOC.25,28 Based on 100 simulated SOCs accomplished with the detailed morphometric analysis, we drew a preliminary conclusion that the supraorbital keyhole approach via an eyebrow incision may be applied to obtain the OB as a source of OECs in ca. 60% of cases. In the current paper we present the results of 17 cadaver dissections performed in order to verify the usefulness of the standard SOC in approaching the OG and obtaining the undamaged OB with the combination of a classic microsurgical and neuroendoscopy-assisted techniques. The cadaver dissection laboratory simulated the environment of an operating theater and the study itself has been designed according to the advice and tips presented in the literature.37 We used fresh cadavers for performing simulated OB harvesting prior to routine autopsy. The brain vessels were not injected with latex or any other contrast solution, which might be assumed as a major drawback of the study. In fact, it decreases the legibility of the anatomical structures but made the conditions similar to those known from the operating field. The extensive search of the literature available did not provide us with any tool allowing the objective assessment of the surgeon’s comfort while performing an operation. Hence, we decided to introduce a simple ten-grade scale—similar to Visual Analog Scale—routinely used for assessing pain. Pain is also highly subjective and somehow irreproducible parameter with the high inter-subject variability.33 The mean operator’s comfort at 80% of the maximum and the average operation time oscillating at 70 min for plateau of the learning curve were acceptable as they seem to reflect the reality.27,28 The main concern would be that the cadavers give no indication of extent of bleeding, which might potentially

prolong the operation. However, we presumed that the time for postoperative endoscopic inspection of OB remnants and neural tissue or vessel injury reflects the time needed for the intraoperative hemostasis. Our results show that it might be possible to obtain an intact OB in roughly 60% of cases which is consistent with our previous radiological observations.25 Surprisingly, we discovered in another 30% cases that the iatrogenic injury of the OB, even though present, not necessarily exclude it for processing for OEC culture. This finding would not be possible without a cadaveric simulation giving the possibility of macroscopic inspection of the harvested OB. Certainly an ideal method of the verification would be the further tissue processing and the establishment of a cell culture, but this was not the case due to the long postmortem time of ischemia, making any attempt for OEC culture impossible.18 Cadaveric tissue at these postmortem intervals is not suitable for culture, and hence culture was not carried out. It would not give useful information. The purpose of the paper was solely to describe a surgical approach. Estimation of the yield and purity of cells cultured from the human OB will require a study of cultures of fresh biopsy material from living patients. We will be reporting results from such an ongoing study separately. The fact that deserves a separate comment is the relatively high—over 40%—incidence of potential basal frontal lobes injuries, reflected by tear of the arachnoid, noted in our series. This may be mainly due to postmortem brain swelling resulting in increased susceptibility to mechanical damage.38,39 Nevertheless, one may state that for cases in which postsimulation neural tissue injury was spotted there was a contact of neurosurgical instruments with the frontal lobes. This may be potentially related to the damage of the neural tissue and should be avoided as a principle. In real life it is not possible to completely avoid minor manipulations on the delicate neural tissue, particularly while creating the approach to the desired structure. Particularly worthy of notice is the fact that in all of reported cases the “injury” was made while approaching optic and carotid cisterns during the first stage of the simulated operation. In this part of the procedure, surgeon needs to advance brain retractor or bipolar forceps down to the basal cisterns sliding in on the base of the frontal lobe. This might lead to the reported tears of the arachnoid that are not normally observed on living individuals. Importantly, this stage is common for all microsurgical procedures performed using keyhole SOC regardless of the exact surgical target (which may be, for example, tuberculum sellae meningioma or aneurysm of the middle cerebral artery). What is really important from the perspective of our study is the fact that none of the reported tears of

Br J Neurosurg Downloaded from informahealthcare.com by Nyu Medical Center on 04/24/15 For personal use only.

8

M. Czyz et al.

the arachnoid were observed in the region of the OG, thus none was directly related to the manipulations in the region of interest. The thesis may be also supported by correlation between the neural tissue injury and two factors: the lowering of the operator’s comfort and the incidence of OB damage which may reflect the worsening of maneuverability in the operating field caused by a generalized tissue fragility. In addition, no vascular injury was discovered which might be assumed as another argument for stating that the real incidence of the intraoperative clinically significant frontal lobe tearing might be lower than that found in our cadaveric series. Interestingly, we discovered a strong negative correlation between the craniotomy width and an event of the neural tissue injury. This observation supports the postulates of the keyhole concept especially when taking into account that the size and location of craniotomies performed by us coincide with the data from the literature.28,31,32 Another approach-related complication observed was the unintended opening of the frontal sinus in 18% of cases, which incidence correlated positively with the craniotomy width bigger than 30 mm. The possibly dangerous complication—due to the risk of nasal CSF leak resulting in meningitis26,27—would have been probably avoided in majority of situations by the routine use of the intraoperative image guidance. Because of technical limitations it was not possible to directly assess the ratio of CSF leak in our cadaveric study. Verifying the usefulness of the preoperative 3D planning, we found significant correlation between the unfavorable skull base configuration and a need for an extensive skull base drilling and unavoidable OB damage. This may be carefully assumed as a proof of reliability of our method of the preoperative virtual preassessment. It may predict technical difficulties and even lack of the possibility of OB harvesting via SOC, which is extremely valuable information for both the patient as well as the surgical and laboratory team. The main drawback of our study was the relatively small number and high average age of cases that underwent simulated microsurgical procedure. However, given that the ongoing and widely known problems with the access to the cadaveric material, one can assume our series to be even more substantial than the average known from the literature.29,40 This allows us to state that our results can be carefully taken into account while planning the approach to an intact OB as a source of OECs.

Conclusions The keyhole SOC via an eyebrow incision may be safely and effectively applied to obtain the intact OB in roughly 60% and to obtain injured but probably still applicable for neuroregeneration OB in another 30% of cases. Virtual 3D planning may be useful in qualification and preoperative assessment of potential donors.

Acknowledgements Authors are thankful to Dr. Alicia Calvo, Faculty of Medicine, University of Barcelona, Spain and Dr. Daniel Gheek, Faculty of Medicine, Wroclaw Medical University, Wroclaw,

Poland for the technical support during cadaveric specimen dissections. Declaration of interest: The authors report no declarations of interest. The authors alone are responsible for the content and writing of the paper. The research has been supported by the National Ministry of Science and Education grant Pbmn-98 in years 2012–2013.

References 1. Li Y, Li D, Raisman G. Interaction of olfactory ensheathing cells with astrocytes may be the key to repair of tract injuries in the spinal cord: the ‘pathway hypothesis’. J Neurocytol 2005;34:343–51. 2. Raisman G. Specialized neuroglial arrangement may explain the capacity of vomeronasal axons to reinnervate central neurons. Neuroscience 1985;14:237–54. 3. Ramon-Cueto A , Munoz-Quiles C. Clinical application of adult olfactory bulb ensheathing glia for nervous system repair. Exp Neurol 2011;229:181–94. 4. Santos-Benito FF, Ramon-Cueto A . Olfactory ensheathing glia transplantation: a therapy to promote repair in the mammalian central nervous system. Anat Rec B New Anat 2003;271:77–85. 5. Kachramanoglou C, Li D, Andrews P, et al. Novel strategies in brachial plexus repair after traumatic avulsion. Br J Neurosurg 2011;25:16–27. 6. Granger N, Blamires H, Franklin RJ, Jeffery ND. Autologous olfactory mucosal cell transplants in clinical spinal cord injury: a randomized double-blinded trial in a canine translational model. Brain 2012;135:3227–37. 7. Bartolomei JC, Greer CA . Olfactory ensheathing cells: bridging the gap in spinal cord injury. Neurosurgery 2000;47:1057–69. 8. Tabakow P, Jarmundowicz W, Czapiga B, et al. Transplantation of autologous olfactory ensheathing cells in complete human spinal cord injury. Cell Transplant 2013;22:1591–612. 9. Li Y, Decherchi P, Raisman G. Transplantation of olfactory ensheathing cells into spinal cord lesions restores breathing and climbing. J Neurosci 2003;23:727–31. 10. Ramon-Cueto A , Cordero MI, Santos-Benito FF, Avila J. Functional recovery of paraplegic rats and motor axon regeneration in their spinal cords by olfactory ensheathing glia. Neuron 2000;25:425–35. 11. Choi D, Law S, Raisman G, Li D. Olfactory ensheathing cells in the nasal mucosa of the rat and human. Br J Neurosurg 2008;22:301–2. 12. Lima C, Pratas-Vital J, Escada P, et al. Olfactory mucosa autografts in human spinal cord injury: a pilot clinical study. J Spinal Cord Med 2006;29:191–203. 13. Kachramanoglou C, Law S, Andrews P, Li D, Choi D. Culture of olfactory ensheathing cells for central nerve repair: the limitations and potential of endoscopic olfactory mucosal biopsy. Neurosurgery 2013;72:170–8. 14. Feron F, Perry C, Cochrane J, et al. Autologous olfactory ensheathing cell transplantation in human spinal cord injury. Brain 2005;128:2951–60. 15. Mackay-Sim A , Feron F, Cochrane J, et al. Autologous olfactory ensheathing cell transplantation in human paraplegia: a 3-year clinical trial. Brain 2008;131:2376–86. 16. Kueh JL, Raisman G, Li Y, Stevens R, Li D. Comparison of bulbar and mucosal olfactory ensheathing cells using FACS and simultaneous antigenic bivariate cell cycle analysis. Glia 2011;59:1658–71. 17. Yamamoto M, Raisman G, Li D, Li Y. Transplanted olfactory mucosal cells restore paw reaching function without regeneration of severed corticospinal tract fibres across the lesion. Brain Res 2009;1303:26–31. 18. Miedzybrodzki R, Tabakow P, Fortuna W, Czapiga B, Jarmundowicz W. The olfactory bulb and olfactory mucosa obtained from human cadaver donors as a source of olfactory ensheathing cells. Glia 2006;54:557–65. 19. Dlouhy BJ, Awe O, Rao RC, Kirby PA , Hitchon PW. Autograft-derived spinal cord mass following olfactory mucosal cell transplantation in a spinal cord injury patient. J Neurosurg Spine 2014;21:1–5. 20. Ransom ER, Chiu AG. Prevention and management of complications in intracranial endoscopic skull base surgery. Otolaryngol Clin North Am 2010;43:875–95.

Br J Neurosurg Downloaded from informahealthcare.com by Nyu Medical Center on 04/24/15 For personal use only.

Obtaining OEC via supraorbital craniotomy 9 21. Ivan ME, Jahangiri A , El-Sayed IH, Aghi MK . Minimally invasive approaches to the anterior skull base. Neurosurg Clin N Am 2013;24:19–37. 22. Kryzanski JT, Annino DJ, Jr., Heilman CB. Complication avoidance in the treatment of malignant tumors of the skull base. Neurosurg Focus 2002;12:e11. 23. Romani R, Lehecka M, Gaal E, et al. Lateral supraorbital approach applied to olfactory groove meningiomas: experience with 66 consecutive patients. Neurosurgery 2009;65:39–52. 24. El-Bahy K. Validity of the frontolateral approach as a minimally invasive corridor for olfactory groove meningiomas. Acta Neurochir (Wien) 2009;151:1197–205. 25. Czyż M, Tabakow P, Gheek D, et al. The supraorbital keyhole approach via an eyebrow incision applied to obtain the olfactory bulb as a source of olfactory ensheathing cells - radiological feasibility study. Br J Neurosurg 2014;28:234–40. 26. Fischer G, Stadie A , Reisch R, et al. The keyhole concept in aneurysm surgery: results of the past 20 years. Neurosurgery 2011;68:45–51. 27. Reisch R, Perneczky A . Ten-year experience with the supraorbital subfrontal approach through an eyebrow skin incision. Neurosurgery 2005;57:242–55. 28. Reisch R, Marcus HJ, Hugelshofer M, et al. Patients’ cosmetic satisfaction, pain, and functional outcomes after supraorbital craniotomy through an eyebrow incision. J Neurosurg 2014;121:730–4. 29. Hong WC, Tsai JC, Chang SD, Sorger JM. Robotic skull base surgery via supraorbital keyhole approach: a cadaveric study. Neurosurgery 2013;72:33–8. 30. Komatsu F, Komatsu M, Inoue T, Tschabitscher M. Endoscopic supraorbital extradural approach to the cavernous sinus: a cadaver study. J Neurosurg 2011;114:1331–7. 31. Perneczky A , Reisch R. Keyhole Approaches in Neurosurgery: Volume 1: Concept and Surgical Technique. Wien New York: Springer, 2008.

32. Reisch R, Perneczky A , Filippi R. Surgical technique of the supraorbital key-hole craniotomy. Surg Neurol 2003;59:223–7. 33. Hawker GA , Mian S, Kendzerska T, French M. Measures of adult pain: Visual Analog Scale for Pain (VAS Pain), Numeric Rating Scale for Pain (NRS Pain), McGill Pain Questionnaire (MPQ), Short-Form McGill Pain Questionnaire (SF-MPQ), Chronic Pain Grade Scale (CPGS), Short Form-36 Bodily Pain Scale (SF-36 BPS), and Measure of Intermittent and Constant Osteoarthritis Pain (ICOAP). Arthritis Care Res 2011;63: S240–52. 34. Tabakow P, Raisman G, Fortuna W, et al. Functional regeneration of supraspinal connections in a patient with transected spinal cord following transplantation of bulbar olfactory ensheathing cells with peripheral nerve bridging. Cell Transplant 2014;23:1631–55. 35. Ibrahim A , Li D, Collins A , et al. Comparison of olfactory bulbar and mucosal cultures in a rat rhizotomy model. Cell Transplant 2014;23:1465–70. 36. Barnett SC, Alexander CL, Iwashita Y, et al. Identification of a human olfactory ensheathing cell that can effect transplantmediated remyelination of demyelinated CNS axons. Brain 2000;123:1581–8. 37. Tschabitscher M, Di Ieva A . Practical guidelines for setting up an endoscopic/skull base cadaver laboratory. World Neurosurg 2013;79: S16.e1–7. 38. Takahashi N, Satou C, Higuchi T, et al. Quantitative analysis of brain edema and swelling on early postmortem computed tomography: comparison with antemortem computed tomography. Jpn J Radiol 2010;28:349–54. 39. Winklhofer S, Surer E, Ampanozi G, et al. Post-mortem whole body computed tomography of opioid (heroin and methadone) fatalities: frequent findings and comparison to autopsy. Eur Radiol 2014;24:1276–82. 40. Comert A , Ugur HC, Kahilogullar G, et al. Microsurgical anatomy for intraoperative preservation of the olfactory bulb and tract. J Craniofac Surg 2011;22:1080–2.