Perilymph Gusher in Cochlear Implantation Surgery: An Algorithm for Management Authors: Ahmed Abdelmoneim Teaima, Lobna El Fiky, Badr Eldin Mostafa, Mina Fayez Saleeb, Ossama Mustafa Mady Published: 2026 ABSTRACT ABSTRACTObjectivePerilymph leak is considered a major complication of cochlear implantation (CI) surgery and can lead to surgical difficulties and postoperative morbidities. This study is aimed to formulate an algorithm for better management of perilymph leak in cases of CI.MethodsThis is a retrospective study conducted in our tertiary institute on pediatric patients with perilymph leak during CI surgery from January 2017 to January 2023. Standard facial recess approach and round window cochleostomy were used in all cases. The clinical data, imaging, details of surgery, management of perilymph leak, and outcome were revised.ResultsOne hundred and fourteen cases of perilymph leak out of 1035 cases of CI surgery (11.1%) were collected. In 108 cases, an oozer was encountered, while in 6 cases a severe gusher occurred. Their ages ranged between 2 and 15 years, with 74 male patients. Our algorithm for control of the leak during surgery included elevation of the head, hyperventilation, mannitol, steroids, and wait until the flow of the gusher decreases. Widening of the cochleostomy, use of an electrode with cork, and muscle packing were also helpful in control of the leak. In severe cases, subtotal petrosectomy was adopted with fat obliteration. All cases were managed properly with our algorithm. Only two cases had persistent leak postoperatively and were managed conservatively without need for revision surgery nor lumbar drain.ConclusionIn CI surgery, proper prediction, diagnosis, and systematic management of perilymph leak including nonsurgical and surgical methods are crucial for better and safe outcomes.Level of EvidenceLevel III. Introduction Cochlear implantation (CI) surgery is considered a safe surgery with a low complication rate. The complications are either surgical or device‐related. Major complications need a longer hospital stay or surgical intervention. Perilymph gusher is an uncommon occurrence during CI. It is considered a major surgical challenge and is frequently seen in cases of inner ear malformations. It has been attributed to communication between the cochlea and the internal auditory canal (IAC) causing an abnormal connection between the perilymphatic and subarachnoid spaces. It may lead to meningitis and its complications [1, 2]. Control of perilymph leak in CI surgery includes variety of strategies in the literature with different outcomes. So, we conducted this study to postulate our algorithm for better management of perilymph leak in cases of CI surgery. Methods This is a retrospective study conducted in our tertiary referral center on pediatric patients scheduled for CI surgery and with perilymph leak. This study included all pediatric patients in the period from January 2017 to January 2023. Exclusion criteria were patients older than 18 years old, lack of follow‐up, and no perilymph leak in CI surgery. Standard facial recess approach and extended round window (RW) cochleostomy were used in all cases. We revised their clinical data, imaging, details of CI surgery, management of perilymph leak, and outcome. We then formulated an algorithm for the proper management of perilymph leak. Results Among the 1035 reviewed cases, there were 114 cases with perilymph leak (11.1%). Their ages ranged between 2 and 15 years with 74 male patients. There was a lamina cribrosa defect in the imaging of six cases who were presented by gusher (Figure 1), while the others (108 cases) were presented by oozer. All cases of perilymph leak were associated with inner ear malformations (IEM). CT temporal bone of IP‐Ⅲ showed absence of modiolus . CT, computed tomography; IP, incomplete partition. In our study, we noticed that perilymph leak cases could be differentiated into perilymph oozer and gusher. Oozer is a minimal leak which lasts less than 10 s, flow of perilymph doesn't exceed the middle ear space, doesn't interfere with electrode insertion and stops spontaneously. While perilymph gusher is profuse and continuous, fills the whole surgical field including middle ear and mastoid spaces, interferes with electrode insertion, and doesn't stop spontaneously. There were 108 cases with oozer and 6 cases with gusher. All cases had IEM. Oozer cases were encountered with wide vestibular aqueduct (67 cases), incomplete partition (IP)‐Ⅱ(39 cases), IP‐Ⅰ(2 cases), while gusher cases were found in association with IP‐Ⅰ(3 cases), IP‐Ⅲ(3 cases, Figure 2). IP‐Ⅲ was significantly correlated with the incidence of gusher than other types of IEM (p < 0.05). Perilymph leak cases distribution in our study. IP, incomplete partition; WVA, wide vestibular aqueduct. All malformations were diagnosed preoperatively by imaging. In cases of suspicion of gusher, the parents were counseled regarding difficulty of operation, hazards of infection, and need for vaccination. As a principle, the surgeon should not leave the surgical theater without completely controlling the cerebrospinal fluid (CSF) leakage around the electrode. We followed a specific algorithm for management of such cases. First, we use non‐surgical measures in conjunction with the anesthesiologist including elevation of the patient's head, hyperventilation, use of mannitol and steroids (Figure 3). We wait 20 min after these non‐surgical measures. If leak continues, we proceed to next step. Algorithm for intraoperative perilymph leak management. When the gusher isn't controlled by non‐surgical measures, we widen the cochleostomy opening, use an electrode with a stopper, and pack the cochleostomy with pieces of muscle. We wait for 10 min to be sure leak is under control. If the flow is not controlled by all the above measures, we shift to fat obliteration of middle ear and mastoid cavity and closure of the eustachian tube by a subtotal petrosectomy. Two cases developed postoperative leak and were controlled by conservative measures for 3 to 5 days to decrease intracranial pressure, including head up position, steroids, and mannitol. In our series, we did not need to use a lumbar drain. One case had the leak coming from the incision postoperatively, so continued in conservative management with wound compression postoperatively, and the leak stopped after 2 days (Figure 4A). One case also developed CSF accumulation under skin in the postauricular and infra‐auricular region which was managed conservatively by compression, head‐up position, steroids, and mannitol and it resolved after 5 days (Figure 4B). A case of persistent cerebrospinal fluid (CSF) gusher from skin incision postoperatively in IP‐Ⅲ(A) and a case of subcutaneous CSF accumulation postoperatively in IP‐Ⅰ with CSF gusher (B). IP, incomplete partition. In all our cases of perilymph leak or IEM, we used C‐arm x‐ray before skin closure to ensure proper position of the electrode. Electrode insertion was properly done in all cases except one case of IP‐Ⅲ. In this case, electrode was misplaced in the IAC and was withdrawn and reinserted properly (Figure 5). Transorbital x‐ray view (Stenvers view) showing misplacement of the electrode into the internal auditory canal (A), then replacement again in the cochlea (B) in a case of IP‐Ⅲ with cerebrospinal fluid gusher. IP, incomplete partition. Discussion Perilymph gusher is considered a major challenge in CI surgery. Perilymph gusher is associated with widening of the lamina cribrosa of the cochlea, leading to profuse CSF leakage during CI due to abnormal communication between perilymphatic and subarachnoid spaces. Cochleostomy will suddenly decrease pressure in the perilymph space, so this will lead to CSF influx into inner ear [3, 4, 5]. The incidence of perilymph gusher is quite different across studies due to the variability of the gusher definition. In our study, we found that perilymph leak can be differentiated into oozer and gusher with specific criteria for each. In their series, Li et al. [6] found three cases of CSF gusher (1.15%). They were IP‐malformation and were controlled by muscle packing of the cochleostomy. This fits well with the oozer cases in our series. Kamogashira et al. [7] reported 12 cases of CSF leak and concluded that large diameter of cochlear modiolar base and vestibular aqueduct with absence of modiolus were risk factors for CSF leak. According to Loundon et al. [8], the incidence of perilymph leak was around 4.1% (20/490) of implanted cases while it was 6.7% by Papsin [9]. Incidence of CSF leak was found 0.4% by Daneshi et al [10]. In our study, incidence of perilymph leak was 11.1% of implanted cases. For proper control of perilymph gusher, we must anticipate the presence of IEM as well as probability of gusher occurrence using thin cuts high‐resolution computed tomography of petrous bone. Once alerted, we follow our specific algorithm to properly manage such cases. This protocol can be applied in all cases of perilymph leak not only IEM cases although most cases of perilymph leak have IEM. This protocol proved to be a successful protocol for control of the leak in all cases in our study intraoperatively. Multicenter series including IEM and non‐IEM cases are recommended to confirm more about this protocol efficiency. Two cases in our series had leak postoperatively and were managed conservatively for 5 days and the leak was stopped from the third day without need for revision surgery nor lumbar drain. This conservative management included hospitalization, head elevation, wound compression, systemic antibiotics, mannitol, and steroids. Patients were discharged safely on the fifth postoperative day. Stankovic et al. [11] recommended performing lumbar drain before cochleostomy and to remain in place for 3 days postoperatively. While Kim et al. [12] suggested large cochleostomy for better packing of expected gusher. Loundon et al. [8] recommended intravenous hyperosmolar treatment during CI surgery in cases of inner ear malformation to decrease the possibility of gusher. Wootten et al. [13] recommended reverse Trendelenburg position, intraoperative mannitol and postoperative acetazolamide to control the CSF flow. Different materials were used in the literature for packing of the cochleostomy site. Wootten et al. [13] used fascia or periosteum with less tendency for atrophy‐like muscle. Loundon et al. [8] used muscle pieces with bio glue. Eftekharian and Amizadeh [14] used pieces of temporalis fascia. Papsin [10] used fascia pieces with fibrin glue. Sennaroglu [5] packed the cochleostomy round piece of fascia with the electrode passing in its center. Some authors recommended pieces of temporalis muscle and fascia; others recommended packing of the middle ear cleft [13]. In their study, Isaiah et al. [15] controlled five cases of perilymph leak with fascial donut in the cochleostomy around the electrode. In their study, Sun and Sun [16] implanted six cases of IP‐Ⅲ, five were done through extended RW approach and one case through RW approach, in which the electrode had gone into the IAC. CSF gusher in all cases was controlled by temporalis fascia packing with no postoperative complications. Audiological outcomes were satisfactory. In their study, Kim et al. [17] found 15 cases of CSF leak out of 72 cases of inner ear anomalies cases underwent CI surgery. They used muscle plugs around the electrode in all cases and mastoid obliteration in three cases with no usage of lumbar drain. There was no postoperative CSF leak in all cases. One case developed meningitis 7 months after the surgery and recovered with antibiotics without neurological sequelae. Vashist and Singh [18] reported nine cases of CSF leak out of 146 cases underwent CI surgery. One case was just oozer and stopped spontaneously, other eight cases were gusher. Four cases did not have any predictable factors for gusher with normal inner ear appearance on radiology. All cases were controlled by packing with periosteum in the RW without lumbar drain nor postoperative leak. They close the wound tightly in three layers with pressure bandage over the mastoid for 24 h. Karatas [19] reported 14 cases of CSF leak out of 307 implanted cases. Control of leak was done by packing with muscle or fat. Lumbar puncture was done in three cases to decrease intraoperative increased CSF pressure. Revision surgery with fat obliteration was done in three cases due to persistent postoperative leakage. They recommended packing of cochleostomy site with fat tissue, which was found to be better than muscle. Adunka et al. [20] found 30 cases presented with CSF leak during CI surgery out of 70 cases of IEM. Leak was controlled by cochleostomy packing in all cases with no need for a lumbar drain. They concluded that perilymph gusher did not affect auditory outcome in hypoplastic or incomplete partition cases (Table 1). Table 1Perilymph leak in the literature.StudyCasesOutcomesLi et al. [6]Three cases of gusherGusher was controlled by temporalis fascia packingKamogashira et al. [7]Twelve cases of leak (3.2%)ControlledEftekharian et al. [14]Fifteen cases of leak (2.87%)Controlled by muscle packingIsaiah et al. [15]Five cases of leakLeak was controlled by muscle packingSun et al. [16]Six cases of gusherGusher was controlled by muscle packing. Electrode in one case went into the IACKim et al. [17]Fifteen cases of leak (3.5%)Leak was controlled by soft tissue packing, while three cases necessitated mastoid obliterationVashist et al. [18]Nine cases of leak (5%)Leak was controlled by soft tissue packingKaratas et al. [19]Fourteen cases of leak (4.56%)Leak was controlled by muscle and fat packingAdunka et al. [20]Thirty cases of leakLeak was controlled by soft tissue packingAbbreviation: IAC, internal auditory canal.John Wiley & Sons, Ltd. Perilymph leak in the literature. Abbreviation: IAC, internal auditory canal. In our study, we used different types of electrodes in cases of perilymph oozer, but in cases of gusher we used an electrode with cork. Proper position of electrode inside the cochlea was always verified by C‐arm and corrected immediately if needed. Electrode selection in IEM depended on type of malformation and availability of the implant. We used shorter electrodes with cork in cases of severe malformation and when expecting gusher as IP‐Ⅰand IP‐Ⅲ. We used straight and precurved electrodes in other cases, but we avoided modiolar electrodes in incomplete partition cases. Wootten et al. [13] documented malposition of electrode in IP‐Ⅲcases as the electrode entered the IAC and reached cerebellopontine angle. Dietz et al. [21] reported 6% rate of electrode migration in cases of CSF leak as risk factor. Heman‐Ackah et al. [22] documented two cases of IPIII where the electrode was in IAC. To avoid this complication, intraoperative 3D volume tomography or fluoroscopic imaging guidance were introduced in some CI centers [3]. In our study, we used C‐arm x‐ray in all perilymph leak cases. Electrode in IP‐Ⅲcase was misplaced in the IAC and was reinserted properly. Conclusion In CI surgery, proper prediction, diagnosis, and systematic management of perilymph leak including nonsurgical and surgical methods are crucial for better and safe outcomes. Author Contributions Ahmed Abdelmoneim Teaima: study design, write the manuscript, submission of the article. Lobna El Fiky: study design, review the manuscript. Badr Eldin Mostafa: collect the data, analysis of the data, review the manuscript; Mina Fayez Saleeb: collect the data, analysis of the data, review the literature; Ossama Mustafa Mady: collect the data, analysis of the data, review the literature. Funding The authors have nothing to report. Ethics Statement The study was done in accordance with the ethical standards the 1964 Helsinki declaration and its later amendments or comparable ethical standards. This study was approved by Ain Shams University, Faculty of Medicine institutional review board FMASU R267/2023. All patients signed informed written consent prior study started. All patients' data were dealt with complete confidentiality, and no one has the right to read their medical information except the main researcher. Conflicts of Interest Authors have no conflict of interest to disclose. Professor Lobna El Fiky is a member of the World Journal of Otorhinolaryngology—Head and Neck Surgery (WJOHNS) editorial board and is not involved in the peer review process of this article.