Clinical and radiological outcomes of arachnoid-preseving suboccipital decompression for adult chiari I malformation with and without syringomyelia

(1)

Contents lists available atScienceDirect

Clinical Neurology and Neurosurgery

journal homepage:www.elsevier.com/locate/clineuro

Clinical and radiological outcomes of arachnoid-preseving suboccipital

decompression for adult chiari I malformation with and without

syringomyelia

Rahsan Kemerdere

a

, Mehmet Yigit Akgun

a

, Semih Can Cetintas

a

, Tibet Kacira

b

,

Taner Tanriverdi

a,

*

a_{Department of Neurosurgery, Cerrahpasa Medical Faculty, Istanbul University-Cerrahpasa, Istanbul, Turkey} b_{Department of Neurosurgery, Medical Faculty, Sakarya University, Sakarya, Turkey}

A R T I C L E I N F O

Keywords: Chiari malformation Craniectomy Duraplasty

Posterior fossa decompression Syringomyelia

A B S T R A C T

Objective: To share our experience related to clinic and radiologic outcomes of patients with Chiari I mal-formation.

Patients and Methods: This retrospective study evaluated surgical outcome of 48 patients who underwent pos-terior fossa decompression and duraplasty with arachnoid-preservation technique February 2010 and February 2019. Clinic and radiologic outcomes at long-term follow-up were provided.

Results: Surgery led to satisfactory outcomes in both clinic and radiologic measures. The majority of patients (66.7%) in the follow-up period informed us that their symptoms signiﬁcantly improved. In 14 patients (29.2%), the symptoms were gone totally and no complaining was reported to us. Syringomyelia was encountered in 21 patients (43.7%) on pre-operative MRI. During long-term, signiﬁcant improvement (syrinx segments were re-markably reduced in diameter) in syrinx size was noted in 16 of the 21 patients (76.1%). In 5 of the 21 patients (23.8), complete improvement, that is, syrinxes were totally collapsed, was seen. Surgery-related complications occurred in a total of 5 patients (10.4%).

Conclusion: In our experience, we found decompressive craniectomy wide enough with duraplasty without arachnoid opening is safe and successful in patients with CIM. Complication rate is low and arachnoid preserving technique should be performed by experienced neurosurgeons.

1. Introduction

The Chiari malformation is a group of disorders including des-cending of the posterior fossa contents below the cranio-cervical junc-tion (CCJ) or foramen magnum and associated anomalies of the sur-rounding bony structures. This entity was first described by the Austrian pathologist Hans Chiari in 1891 [1]. Chiari I malformation (CIM) is the most common form which include downward herniation of the caudal part of the cerebellum below the foramen magnum and can be seen in both children and adults [2,3]. Obstruction of the cere-brospinalfluid (CSF) circulation or dynamics at the CCJ leads to pres-sure gradient between the cranial and spinal regions that finally syr-ingomyelia may develop [4,5]. Clinical manifestations can be diverse resulting from the repercussions of the malformation on the brainstem or spinal cord. The exact cause of this malformation including the

formation of syringomyelia is not known very well but previous sig-niﬁcant studies suggested that underdevelopment of the posterior fossa bony structures may be the main cause because after bony decom-pression, the majority of clinical manifestations resolve [6,7].

The main treatment of CIM is surgery however; there has been no common consensus and there is still a discussion in the current litera-ture related to surgical strategy. We have to respect the Masters [4–6,8] who published surgical series previously that tried toﬁnd out optimal surgical strategy with fewer surgical complications and favorable clin-ical and radiologclin-ical outcomes. There has been only one common no-tion which is the posterior fossa decompression (PFD) among neuro-surgeons since Hans Chiari described the entity [1]. The diﬀerences in surgical techniques come from the adjuvants to PFD. The adjuvants range from minimally traumatic approach such as splitting the outer layer of the dura, to more aggressive approach such as resection of the

https://doi.org/10.1016/j.clineuro.2019.105598

Received 26 September 2019; Received in revised form 7 November 2019; Accepted 11 November 2019 ⁎_{Corresponding author.}

E-mail addresses:rakemerdere@yahoo.com(R. Kemerdere),myigitakgun@gmail.com(M.Y. Akgun),semihcan1992@gmail.com(S.C. Cetintas), tibetkacira@yahoo.com(T. Kacira),tanerato2000@yahoo.com(T. Tanriverdi).

Available online 12 November 2019

(2)

cerebellar tonsils and syringo-subarachnoid shunts. The most common surgical technique performed in CIM is the combination of PFD and opening of the arachnoid and dura then duraplasty [9]. It is clear that there has also been no common consensus on this technique that debate still continues about the size of the posterior fossa craniectomy, pre-serving the arachnoid and even use of duraplasty materials in surgical series and even in comparative studies. Studies in which PFD plus arachnoid-preserving duraplasty are limited in the current literature [8,10–12] and our aim is to share our experience in a series of adult patients with CIM in which PFD plus duraplasty with arachnoid-pre-serving surgical technique was performed.

2. Patients and methods 2.1. Patients and data collection

This retrospective analysis retrieved 53 consecutive patients who have undergone surgery on CIM between February 2010 and February 2019. This study was approved by the institutional review board. The following inclusion criteria were applied: patients who were (i) an age of at least 18 years old at the time of surgery; (ii) diagnosed with CIM; (iii) had no other cranial and spinal pathologies and no previous cranial surgery; and (iv) had a minimum of 6 months of post-operative follow-up. Depending on the selection criteria,ﬁnally a total of 48 patients were included. All patients were operated by the same surgeon (TT) with the same surgical technique after the signed informed consents were obtained either from the patients or next of kin.

2.2. Pre-operative work-up and data assessment

Demographic variables, clinical signs and symptoms were carefully noted in every patient. All patients provided head and spinal magnetic resonance imaging (MRI) studies in order to see presence or absence of hydrocephalus and syringomyelia, respectively. Position of the pos-terior fossa contents, especially cerebellar tonsils with the reference to basion-opisthion line and if present, location and extension of syr-ingomyelia were carefully noted. All patients were informed related to the complications that may be seen during or after surgery.

2.3. Surgical approach

All symptomatic patients with cerebellar tonsil herniation with or without syringomyelia underwent surgery. After getting prone position, surgery started with a midline skin incision from 1 cm above the inion to C2. After passing through the skin, nuchal ligament was dissected and a piece wide and long enough was removed for duraplasty. Para-spinal muscles were dissected in sub-periosteal fashion. After hemos-tasis, two burr-holes on each side of the midline were performed on the sub-occipital bone. The craniectomy was started by combining the two burr-holes with kerrison rongeur. The size of our craniectomy was in general 4 × 4 cm or 4 × 3 cm: 2 cm from the midline on each side and 3 or 4 cm height from the CCJ. We never had craniectomy condyle to condyle and never saw sigmoid sinus laterally and transverse sinus superiorly. After the craniectomy wasﬁnished, C1 laminectomy (1 cm on each side of the midline) and removal of a small piece of bone from C2 lamina was done. The aim of this small bone removal from C2 was to perform easy or comfortable duraplasty. Total C2 laminectomy was only preferred in cases in which the tonsillar herniation reached C2 or below. Following bone removals, dura was inspected under the mi-croscope and a thickened band of dura was found at the CCJ in all patients. The thickened band of dura was dissected and removed at the CCJ before opening the dura. The dura incision was started just in front of C2 on the midline and vertically extends to the CCJ. As soon as the incision passed the CCJ, the incision extended in curvilinear fashion toward the right or left cerebellar hemisphere (no criteria for selecting the right or left side) irrespective of severity of tonsillar herniation.

Dura incision was performed carefully with the intent to preserve the arachnoid layer. We never used“Y-shaped” dura incision and the most advantageous part of our incision was that we never encountered oc-cipital sinus located midline and it is very easy and safe to perform. In every case, after opening the dura, we had expansion of the arachnoid layer at the CCJ showing decompression. Under the microscope, sub-arachnoid space was inspected through the intact sub-arachnoid and the tip of the herniated tonsils was seen. If the arachnoid layer was inad-vertently opened, we sealed it by using very-low current bipolar elec-trocautery. Duraplasty with nuchal ligament graft (we never used periosteum or any synthetic material) was performed in watertight fashion with 3-0 coated vicryl®. After the completion of duraplasty, a Valsalva maneuver was performed 3 times and held at 60 mmHg for 5 s to check out any CSF leakage. After we sure that there was no CSF leakage following Valsalva maneuver, dural closure was augmented withﬁbrin glue.

2.4. Post-operative care and follow-up

On average, patients were discharged 5 days after surgery and the first head and spinal MRI was performed at 3 months. The second follow-up MRI withflow sequence was done at 12 months. In patients who had no syringomyelia on pre-operative MRI and without any complaining, we did not perform MRI anymore after 12 months and they were clinically checked annually. In patient who had syr-ingomyelia on pre-operative MRI, we perform head and spinal MRI annually until we were sure that syringomyelia was completely re-solved. Thus, all patients with or without syringomyelia on pre-opera-tive MRI and without complaining, were clinically checked annually. If patients were not able to come to see us, we reached by telephone to get information about their clinical situation. For this study, clinical and radiological outcomes were measured depending on our objective findings, patients’ overall self-rated feelings and head and spinal MRI were evaluated by our radiology departments which had no informa-tion about the present study. Clinical outcome was evaluated as follows: “worse”, “stable”, “significantly improved”, and “completely im-proved”. Radiological evaluations, especially size of syringomyelia were as follows:“worse” means increase in size, “stable” means no change in size”, “significantly improved” means the syrinx was almost resolved, and“completely improved” means the syrinx was totally re-solved or no syrinx.

2.5. Statistical analysis

We used a commercially available statistical software package (SPSS version 22.0; IBM Corporation, Inc., Chicago, IL, USA) for all the sta-tistical analyses. The mean ± standard deviations ( ± SD) were cal-culated for each parameter. Chi-square tests were used for appropriate comparisons. Diﬀerences were considered statistically signiﬁcant if the probability value was less than 0.05.

3. Results

3.1. Pre-operative clinical and radiological measures

This study included 48 patients with a mean age of 38.8 ± 12.9 years (range 18–69 years). We noted a significant female predominance with 70.8% against 29.2% in males and the difference was significant (p = 0.004). Mean symptom duration was 21.5 ± 28.3 months with a range 6–120 months. All patients were symptomatic and the most common presenting symptom was sub-occipital headache (72.9%) which was increased by a Valsalva maneuver in all patients. In some, sub-occipital headache was together with neck pain (Table 1). Several other symptoms in addition to headache were noted (Table 2). Among the additional symptoms, sensory disturbances were the most common (20.8%) followed by imbalance and gait disturbances. Our neurological

(3)

examinations showed that the majority of patients (72.9%) were without neurological deficits. Interestingly the most common abnormal clinical finding was decreased gag reflex (10.4%) followed by hor-izontal nystagmus (4.2%). Several other neurological examination findings are listed inTable 3.

Radiologically, all patients had tonsillar herniation more than 5 mm and in 21 patients (43.8%) we noted syringomyelia in which cervical region was the most common location. The majority of syringomyelia was pan-medullary.

3.2. Surgery and surgery-related complications

All patients underwent sub-occipital craniectomy, C1 laminectomy and duraplasty with preservation of the arachnoid layer. In one patient, we had to perform an additional C2 partial laminectomy because of incomplete tonsillar decompression after C1 laminectomy. Our cra-niectomy measured 4 × 4 or 4 × 3 cm (4 cm wide and 3 or 4 cm height) and we never performed wide craniectomy measuring condyle to con-dyle. We never reached laterally to see the sigmoid sinus and superiorly to see the transverse sinus. For duraplasty, we never harvested perios-teum by performing additional skin incision. Instead, we had nuchal ligament harvested after midline skin incision reached to the ligament and additional skin incision was not needed. During dura opening, meticulous care was given not to open the arachnoid layer. In only 2

patients, accidental pin-holing of the arachnoid occurred. In one the pinhole was small and low-voltage bipolar cauterization was applied and in other one accidental arachnoid tear was a little bit larger, then, the arachnoid layer had to be opened fully.

Surgery-related complications (Table 4) occurred in a total of 5 patients (10.4%). Superﬁcial wound infection was seen in 2 patients (4.2%): one required surgical debridement followed by antibiotics and the other one was treated by antibiotic only and both did well during follow-up period. Two patients in whom inadvertent arachnoid tear had occurred during surgery had CSF leakage. One required surgery in which additional more stitches were applied on duraplasty and during follow-up the patient did well. Other one did not require surgery but after we strengthened the skin incision with additional stitches, the leakage resolved spontaneously and the patient did well during follow-up. One patient underwent ventriculo-peritoneal shunting due to the development of communicating hydrocephalus almost one month after surgery and no clinical problem was encountered during follow-up period. In this cohort there were no immediate peri-operative medical or surgery-related complications and there was no death and long-term surgical complication.

3.3. Follow-up: clinic and radiologic outcomes

Table 4 summarizes clinic and radiologic outcomes. The mean follow-up was 2.51 ± 2.59 years with a range of 6 months to 9 years. The majority of patients (66.7%) in the follow-up period informed us that their symptoms significantly improved. They were complaining of slight headache and verbalized that the headache was not a big issue in their life because it occurs sometimes and resolves without medication. In 14 patients (29.2%), the symptoms were gone totally and no com-plaining was reported to us. Only 2 patients, one with syringomyelia on pre-operative MRI, informed us about their neck stiffness which was severe and we had to prescribe medication. During follow-up, every effort was made to treat neck stiffness but we failed. In their follow-up MRI, we noted no pseudo-meningocele and tonsils were in their normal position and the syringomyelia was significantly improved.

Before surgery, abnormal neurological finding was found in 13 (27%) and neurological examination disclosed nothing abnormal in 35 (72.9%) patients. Our neurological examinations showed that abnormal neurological findings significantly improved in 6 (46.1%) and com-pletely improved or neurologic status became normal in 5 (38.4%) patients. Two patients who had no syringomyelia on pre-operative MRI showed that the abnormalfindings were stable. In one of the 2 patient with stable neurologicalfinding, we noted that hypoesthesia on the left upper extremity still continued. In other one, decrease in gag reflex Table 1

Main presenting symptoms and some demographic data of the 48 patients with Chiari I malformation before surgery.

Parameters Number (%)

Male/Female 14/34 (29.2/70.8)

Mean age at surgery (years) 38.8 ± 12.9 Mean symptom duration (month) 21.5 ± 28.3

Head pain 35 (72.9)

Neck pain 4 (8.3)

Head and neck pain 6 (12.5)

Table 2

Other symptoms in addition to head and neck pain in the 48 patients with Chiari I malformation before surgery.

Sensory disturbances 10 (20.8) Imbalance and gait disturbances 7 (14.6)

Motor deﬁcit 4 (8.3)

Dizziness 4 (8.3)

Dysphagia 1 (2.1)

Motor deﬁcit and imbalance 1 (2.1) Dizziness and imbalance 1 (2.1) Sensory disturbance and imbalance 1 (2.1) Dysphagia and imbalance 1 (2.1)

Hearing disorders 1 (2.1)

Tinnitus 1 (2.1)

Table 3

Neurological examination ﬁndings in 48 patients with Chiari I mal-formation before surgery.

None 35 (72.9)

Decreased gag (pharyngeal) reﬂex 5 (10.4)

Nystagmus 2 (4.2) Facial paralysis 1 (2.1) Decrease hearing 1 (2.1) Extremity paresis 1 (2.1) Cerebellar syndrome 1 (2.1) Sensory deﬁcits 1 (2.1) Pyramidal syndrome 1 (2.1) Table 4

Clinical and radiological outcomes of 48 patients with Chiari I mal-formation at the last follow-up.

Mean follow-up (years) 2.51 ± 2.59 Clinical outcome (symptoms)

Signiﬁcantly improved 32 (66.7) Completely improved 14 (29.2)

Stable 2 (4.2)

Neurological outcome (signs)

Normal before surgery 35/48 (72.9) Signiﬁcantly improved 6/13 (46.1) Completely improved 5/13 (38.4)

Stable 2/13 (15.3)

Radiological outcome (syringomyelia)

Signiﬁcantly improved 16/21 (76.1) Completely improved 5/21 (23.8) Complications

Superﬁcial wound infection 2 (4.2) Cerebrospinalﬂuid leakage 2 (4.2)

(4)

with the same severity as in pre-operative period continued but patient informed us that this issue did not cause severe problem in his life. The head MRI from both patients was normal.

Syringomyelia was encountered in 21 patients (43.7%) on pre-op-erative MRI. Interestingly we noted decrease in size in all syrinxes on MRI at 3 months after surgery (Fig. 1). During long-term (Fig. 2), sig-niﬁcant improvement (syrinx segments were remarkably reduced in diameter) in syrinx size was noted in 16 of the 21 patients (76.1%). In 5 of the 21 patients (23.8), complete improvement, that is, syrinxes were totally collapsed, was seen. We never encountered a patient whose syringomyelia either increased or stable in size. Finally, a wider cisterna magna and ascent of the tonsils into the posterior fossa with more rounded shape were seen in all patients with or without the presence of syringomyelia (Fig. 3).

4. Discussion

4.1. Clinical and radiological outcomes

It is clear from the current literature that precise evaluation of clinical and radiological outcomes after surgical management of CIM is very diﬃcult because of diverse clinical signs and symptoms and

different surgical techniques used. Several outcome scores have been reported [13–15] but none of them became a validated and generally-accepted score that allows us having common and informative eva-luation outcome. In this report we did not use any score and our results depend on our clinical and radiologicalfindings and patients’ feedback. Our results demonstrated that the improvements in the presenting symptoms and clinical deficits were 95.9 and 84.5%, respectively. More importantly, the rate of resolution of the syringomyelia was found to be 99.9% at the last follow-up. Headache was the most common presenting symptom that responded well to surgery and the symptom that was difficult to treat in our patient group after surgery was neck stiffness (4.1%). The rate of neck stiffness is very low in our patient group compared to others [3,10,11,16] and the major cause of this debili-tating symptom as reported previously may be due to the formation of pseudo-meningocele [11]. The very low rate of neck stiffness in our patient group may be explained by the fact that preserving the ara-chnoid layer decreased risk of CSF leakage. Our rate of favorable out-comes is comparable to the current literature in which symptom im-provement has been reported in the range of 70–100%. Syrinx reduction or complete resolution was reported in the range of 50 to almost 100% [10–12,16–19] but we have to underline that the rate changes depending on the type of surgery. In one analysis [20], it has been reported that improvement in syrinx size was 50% with simple sub-occipital craniectomy only and it was almost 100% if duraplasty was added. There has been a common consensus that favorable clinical and radiological outcomes are higher in duraplasty group compared to non-duraplasty group [3,20]. And no significant differences were noted regarding clinical outcome and syrinx reduction between those who had arachnoid dissection and those who had arachnoid preservation [8]. In limited number of studies in which sub-occipital craniectomy and duraplasty without arachnoid dissection reported previously [8,10–12], the rate of syrinx reduction or complete resolution reached up to 90%. Our results demonstrated that arachnoid-preserving tech-nique is quite successful and syrinx reduction and disappearance were found to be 76.1 and 23.8%, respectively. Thus, overall favorable im-provement in syrinx reduction in the present series is almost 100%. Our findings related to syrinx reduction are comparable with the current literature [8,10–12]. Furthermore, our results showed the tonsils have Fig. 1. Pre-perative T2-weighted sagittal MRI (a) showing tonsillar herniation

and syringomyelia extending to the dorsal medulla. Post-operative T2-weighted sagittal MRI (b) at 3 months after surgery. Note that cerebellar tonsils with rounded tip have ascended to its normal position and signiﬁcant decrease in size of the syrinx cavity.

Fig. 2. Comparing pre-operative (a) and post-operative (b) T2-weighted sagittal MRI shows signiﬁcant decrease in syrinx size and normal morphology of the cerebellar tonsils at long-term follow-up (1 year after surgery).

Fig. 3. Complete resolution of the syrinx was noted in this patient at 6 months after surgery and cerebellar tonsils has regained their normal shape and are in normal anatomical position comparing to preoperative MRI (a; pre-operative and b; post-operative MRI).

(5)

ascended to their normal anatomic position and the tips have become more rounded shape in all patients even at 3 months after surgery. Our results and the current clinical studies again supported the concept that the tendency of the cerebellar tonsils to return to its normal anatomical position and achieving normal shape of the tonsillar tip is because of CIM is a secondary deformation due to underdevelopment of the pos-terior fossa bony structures and not a primary abnormality of the cer-ebellum [6,7,10].

4.2. Surgery-related complications

In patients who underwent surgery for CIM, we see different range of surgery-related complications. Frequency of the same complication shows diverse range in different centers. This difference is under-standable because the incidence of postoperative complication is in-fluenced by the surgical techniques used. The range of surgical com-plication in the literature is between 3–40% [10,11,17,21]. Studies indicated that low complication rate is seen if the dura is left intact but in combined technique including dural and arachnoid opening, the rate increased up to 42% [12]. The most common complication has been reported to be meningitis and CSF leakage in patients who had dura and arachnoid opening but in arachnoid preserving group incidence of meningitis and CSF leakage decreased [10–12]. The complication rate in our group was 10.4% which is in line with the studies composed of patients in whom the arachnoid was left intact. We did not have pseudo-meningocele in our patients and CSF leakage and superficial wound infection were the most common. Some authors have reported that non-autologous grafts were associated with more adverse reactions and low complication rate has been reported with autologous graft such as pericardium [22,23]. However, some others reported poor outcome with one of the autologous graft, nuchal ligament and suggested to use pericranium [11]. Interestingly, some group reported that to reduce the risk of meningitis, using autologous materials should be avoided [14]. We did not have a chance to provide difference between autologous graft versus non-autologous graft or between pericranium and nuchal ligament in this report. But depending on our own results, we support to use autologous grafts or in common term, patients’ own tissues (very cheaper or free compared to very expensive non-autologous grafts!!!) in order to minimize graft-related complications. Harvesting the nuchal ligament from within the same midline incision eliminates additional scalp incision for pericranial graft and we did not encounter any ad-verse effect and CSF leakage was seen in only 2 patients (4.2%). Our results are in line with a recent large survey study [24] that preserva-tion of the arachnoid layer decreases the incidence of symptomatic aseptic meningitis and CSF leakage because opening the arachnoid layer exposes the subarachnoid space to cellular debris and blood products.

4.3. Surgical technique

The aim of surgery in CIM is to provide more space for posterior fossa contents by performing bony decompression and additional dur-aplasty. By doing the procedure, surgery creates wide cisterna magna or make artificial cistern which in turn restores CSF flow and normalizes the pressure gradient between cranial and spinal regions. The pathology is confined to the foramen magnum and the area under the arch of C1, hence bony decompression should focus on the midline. Our surgical technique resulted in favorable clinic and radiologic outcomes. Regarding the surgical technique, almost every step is still a matter of discussion. First, there is no consensus about the extent of bone de-compression. Some groups are performing small craniectomy [16,25] but some others have wide craniectomy measuring condyle to condyle [11,17]. It is clear that smaller craniectomy can cause inadequate de-compression and may increase the probability of the second surgery. On the other hand large or wide craniectomy can cause herniation of the cerebellum, which can need cranioplasty. In our series we performed

4 × 4 cm (2 cm from the midline) craniectomy which allowed favorable outcome and no cerebellar herniation. We do not use navigation in order to identify location of the sigmoid and transverse sinuses or their junction which has been used by some authors who performed wide craniectomy condyle to condyle [11]. A more extensive lateral exposure is unnecessary which may expose the patient to unnecessary vertebral artery injury [26]. Whatever the size of craniectomy, we are certainly sure that bony decompression is sine qua non in surgical treatment of CIM. Second, the discussion whether duraplasty should be performed and optimal dural type of plasty, continues. We agree with the authors who suggested duraplasty in order to reduce the risk of some important post-operative complications such as CSF leakage, meningitis, hydro-cephalus and pseudo-meningocele [8,10,13,16]. Furthermore, dur-aplasty could reduce the adhesions of overlying muscles to the ara-chnoid layer. In contrast to some others [11,16], we never used tack-up sutures of the duraplasty to the bony boundaries and to the muscles at the center and we had no problem in none of the patients. Regarding the type of duraplasty, we always used the nuchal ligament harvested from within the midline incision. Some argue against using autologous flaps such as pericranium [14] or nuchal ligament [11] in order to decrease arachnoid adhesion but some others [22,23] like us support autologous grafts because it can cause low complication rate and no inflammatory reactions. Interestingly, poor outcomes have been re-ported by using the nuchal ligament as for duraplasty [11]. We always used the nuchal ligament for duraplasty and only 2 patients showed CSF leakage. We think that there should not be a difference between the pericranium and nuchal ligaments, both of which belong to the pa-tient’s own tissues and in experienced hands nuchal ligaments may be as successful as pericranium for duraplasty. In addition, by harvesting the nuchal ligament from within the same midline skin incision elim-inates additional skin incision for harvesting the pericranium. We agree with the majority of the authors that autologous grafts are more com-fortable to the patients and decrease economic burden because non-autologous grafts are very expensive and the money matters in devel-oping or underdeveloped countries. Third, another debate which still continues is whether the arachnoid layer is opened or not. It is clear that opening the arachnoid exposes the subarachnoid space to cellular debris and blood products which increases risk of subarachnoid adhe-sions or arachnoid adheadhe-sions and aseptic meningitis which has been considered to be a source of failure. Thus, we agree with the recent series [10–12] that opening and the extensive subarachnoid dissection should be avoided because several authors have suggested that ara-chnoid scarring is associated with unfavorable long-term outcome [14,17]. In fact, arachnoid-preserving technique is quite old having been introduced by Logue and Edwards [27] in order to reduce mor-tality rates associated with the original Gardner operation in the 1960′s and 1970′s [4,28]. Gardner [4,28] suggested that CSF is directed to-ward the central canal in the spinal cord to form syrinx by the ob-structing tonsils, thus blocking the mouth of the central canal to restore normal hydrodynamicflow. Gardner plugged the opening of the central canal to the forth ventricle with a piece of muscle. This surgical tech-nique was actually used by Loque and Edwards [27] in some of their cases but then they abandoned the procedure because mainly of a high incidence of complications. Thus, we agree with Loque and Edwards [27] who introduced arachnoid-preserving technique for thefirst time that decompression of the foramen magnum with keeping the ara-chnoid intact to avoid any blood entering the subaraara-chnoid space is safe, effective and decrease the incidence of CSF fistula. As underlined by the authors [27] any neural abnormality can be seen through the arachnoid.

The rate of re-operation for persistent symptoms is comparable in patients those who had extra-arachnoid and those who had intra-ara-chnoid surgical techniques. A recent study has found that there is no signiﬁcant diﬀerence regarding clinical and radiological outcomes be-tween intra-arachnoid dissection with tonsillar manipulation and extra-arachnoid dissection and further intra-extra-arachnoid dissection has caused

(6)

more post-operative complications [12]. According to these authors [12] and some others [14], opening of the arachnoid does not guar-antee a greater eﬀectiveness and intra-arachnoid dissection with ton-sillar manipulation should be reserved for second surgery (rescue pro-cedure) in cases in which inadequate CSF circulation through CCJ is insuﬃcient [11,14]. Depending on their comparative study, they sug-gested that osteo-dural decompression without arachnoid opening is the most appropriate form of treatment of CMI [12]. Our experience showed that extra-arachnoid procedure resulted in favorable outcome and no aseptic meningitis or pseudo-meningocele was seen.

4.4. Study limitations

The retrospective design, limited number of patients and relatively short follow-up are the principal limitations of the present study. Several questions related to surgical technique can only be answered by performing prospective study design with larger patient cohorts in the future.

5. Conclusion

In our experience, we found decompressive craniectomy wide en-ough with duraplasty without arachnoid opening is safe and successful in patients with CIM. Complication rate is low and arachnoid preserving technique should be performed by experienced neurosurgeons. Many issues regarding surgical techniques remain in discussion but for-tunately, almost all diﬀerent surgical techniques have provided favor-able outcome with varying degrees and all made great progress in un-derstanding of this clinical entity.

Funding

This research did not receive any speciﬁc grant from funding agencies in the public, commercial, or not-for-proﬁt sectors.

Declaration of Competing Interest None.

References

[1] H. Chiari, Ueber veranderungen des kleinhirns infolge von Hydrocephalie Des Grosshirns, Dtsch. Med. Wochenscriﬀ. 17 (1891) 1172–1175.

[2] A.J. Barkovich, F.J. Wippold, Sherman J.L, et al., Signiﬁcance of cerebellar tonsillar position on MR, Am. J. Neuroradiol. 7 (1986) 795–799.

[3] J.L. Zhao, M.H. Li, C.L. Wang, et al., A systematic review of Chiari I malformation: techniques and outcomes, World Neurosurg. 88 (2016) 7–14.

[4] W.J. Gardner, Hydrodynamic factors in Dandy-Walker and Arnold-Chiari mal-formations, Childs Brain 3 (1977) 200-121.

[5] B. Williams, Syringomyelia, Neurosurg. Clin. N. Am. 1 (1990) 653–685. [6] T.H. Milhorat, M.W. Chou, E.M. Trinidad, et al., Chiari I malformation redeﬁned:

clinical and radiographicﬁndings for 364 symptomatic patients, Neurosurgery. 44 (1999) 1005–1017.

[7] S. Aydin, H. Hanimoglu, T. Tanriverdi, et al., Chiari type I malformations in adults:

a morphometric analysis of the posterior cranial fossa, Surg. Neurol. 64 (2005) 237–241.

[8] M. Sindou, J. Chavez-Machuca, H. Hashish, et al., Cranio-cervical decompression for Chiari type-I malformation, adding extreme lateral foramen magnum opening and expansile duroplasty with arachnoid preservation. Technique and long-term functional results in 44 consecutive adult cases-comparison with literature data, Acta Neurchir. (Wien) 144 (2002) 1005–1019.

[9] B.G. Rocque, T.M. George, J. Kestle, et al., Treatment practices for Chiari mal-formation type I with syringomyelia: results of a survey of the American Society of Pediatric Neurosurgeons, J. Neurosurg. 8 (2011) 430–437.

[10] H.S. Lee, S.H. Lee, E.S. Kim, et al., Surgical results of arachnoid-preserving posterior fossa decompression for Chiari I malformation with associated syringomyelia, J. Clin. Neurosci. 19 (2002) 557–560.

[11] S. Chotai, V.R. Kshettry, T. Lamki, et al., Surgical outcomes using wide suboccipital decompression for adult Chiari I Malformation with and without syringomyelia, Clin. Neurol. Neurosurg. 120 (2014) 129–135.

[12] C.H.F. Vidal, A.M. Brainer-Lima, M.M. Valenca, et al., Chiari 1 malformation sur-gery: comparing non-violation of the arachnoid versus arachnoid opening and thermocoagulation of the tonsils, World Neurosurg. 121 (2019) E605–E613. [13] L. Aliaga, K.E. Hekma, R. Yassari, et al., A novel scoring system for assessing Chiari

malformation type I treatment outcomes, Neurosurgery. 70 (2012) 656–665. [14] J. Klekamp, Surgical treatment of Chiari 1 malformation-analysis of intraoperative

ﬁndings, complications, and outcome of 371 foramen magnum decompressions, Neurosurgery. 71 (2012) 365–380.

[15] K. Koc, Y. Anik, I. Anik, et al., Chiari 1 malformation with syringomyelia: corre-lation of phase-contrast cine MR imaging and outcome, Turk. Neurosurg. 17 (2007) 183–192.

[16] S.Y. Elhadji Cheikh Ndiaye, L. Troude, M. Al-Falasi, et al., Chiari malformations in adults: a single center surgical experience with special emphasis of clinical im-provement, Neurochirurgie 65 (2019) 69–74.

[17] N. Aghakhani, F. Parker, P. David, et al., Long-term follow-up of Chiari-related syringomyelia in adults: analysis of 157 surgically treated cases, Neurosurgery. 64 (2009) 308–315.

[18] N. Nagoshi, A. Iwanami, Y. Toyama, et al., Factors contibuting to improvement of syringomyelia after foramen magnum decompression for Chiari type I malforma-tion, J. Orthop. Sci. 19 (2014) 418–423.

[19] M.S. Gurbuz, M.Z. Berkma, E. Unal, et al., Foramen magnum decompression and duraplasty is superior to only foramen magnum decompression in Chiari mal-formation type I associated with syringomyelia in adults, Asian Spine J. 9 (2015) 721–727.

[20] S.R. Duraham, K. Fjeld-Olenec, Comparison of posterior fossa decompression with and without duroplasty for the surgical treatment of Chiari malformation type I in pediatric patients: a meta analysis, J. Neurosurg. Pediatr. 2 (2008) 42–49. [21] S.R. Parker, P. Harris, T.J. Cummings, et al., Complications following

decompres-sion of Chiari malformation type I in children: dural graft or sealant? J. Neurosurg. Pediatr. 8 (2011) 177–183.

[22] F.J. Attenello, M.J. McGirt, G.L. Garces-Ambrossi, et al., Suboccipital decompres-sion for Chiari I malformation: outcome comparison of duraplasty with expanded polytetraﬂuoroethylene dural substitute versus pericranial autograft, Childs Nerv. Syst. 25 (2009) 183–190.

[23] A.A. Abla, T. Link, D. Fusco, et al., Comparison of dural grafts in Chiari decom-pression surgery: review of the literature, Spine. 1 (2010) 29–37.

[24] F. Shweikeh, D. Sunjaya, M. Nuno, et al., National trends, complications, and hospital charges in pediatric patients with Chiari malformation type I treated with posterior fossa decompression with and without duraplasty, Pediatr. Neurosurg. 50 (2014) 30–37.

[25] L.Y. Geng, X. Liu, Y.S. Zhang, et al., Dura-splitting versus a combined technique for Chiari malformation type I complicated with syringomyelia, Br. J. Neurosurg. 27 (2018) 1–5.

[26] R.S. Tubbs, M.D. Smyth, J.C. Wellons, Distances from the atlantal segment of the vertebral artery to the midline in children, Pediatr. Neurosurg. 39 (2003) 330–334. [27] V. Logue, M.R. Edwards, Syringomyelia and its surgical treatment-an analysis of 75

patients, J. Neurol. Neurosurg. Psychiatry. 44 (1981) 273–284.

[28] W.J. Gardner, H.S. Bell, P.N. Poolos, et al., Terminal ventriculostomy for syr-ingomyelia, J. Neurosurg. 46 (1977) 609–617.