Endoscopic Ear Surgery for the Management of Traumatic Ossicular Disruption with Intact Tympanic Membrane

Background Traumatic ossicular disruption (TOD) usually had a severe conductive hearing loss, the exploratory tympanotomy is critical for the diagnosis and improve hearing. Endoscopic ear surgery (EES) is becoming popular in the last decade, we conducted a retrospective study to explore the ecacy of EES for management of TOD and the accompanied injuries. Methods A retrospective study was performed on 18 ears (16 patients) of TOD with intact TM from May 2017 to Jun 2019 in our department. EES was conducted to check the ossicular chain anomalies, and to perform the ossiculoplasty and facial nerve (FN) decompression depending on the intraoperative ndings. Hearing outcomes and surgical complications were assessed at 6 months postoperatively. Results The incus injury was the most common type of TOD, which was observed in 14 ears (77.8%), stapes suprastructure fracture was observed in 4 ears (22.2%). FN injury was found in 4 out of 13 ears with temporal bone fracture (TBF), the injury sites were mainly located in the perigeniculate area and the tympanic segment of FN. It showed the postoperative average pure-tone average (PTA) gain was 22.9 ± 9.5 dB, and the average ABG closure was 22.2 ± 8.3 dB, ABG closure to 20 dB or less and ABG closure to 10 dB or less were achieved in 18 ears (100%) and 14 ears (77.8%), respectively. The facial function achieved favorable recovery of House-Brackmann (H-B) grade (cid:0) (3 ears) and grade (cid:0) (1 ear) in all the 4 cases in 6 months after surgery. No iatrogenic FN paralysis and signicant sensorineural hearing loss were observed. injuries, such as it showed favorable surgical outcomes. ESS provides an alternative method with advantage of excellent vision and less


Background
Traumatic ossicular disruption (TOD) is usually caused by a head trauma or direct and indirect injuries from auditory canal [1][2][3].
TOD patients usually had a severe conductive hearing loss, which is induced by tympanic membrane (TM) perforation, ossicular injury, or a combination [3]. Perforations of the TM could naturally heal and then present as an intact one, which is prone to being misdiagnosed [3]. Though high-resolution CT (HRCT) scans and modern imaging technology improve the diagnosis sensitivity before surgery, some TOD cases with the minor abnormalities were still missed in clinical practice [4]. The patients with a persistent severe conductive hearing loss after head or ear trauma, though an intact tympanic membrane upon examination, TOD should be suspected, the exploratory tympanotomy is critical for the diagnosis, and ossicular reconstruction could be performed at the same time according to the intraoperative ndings.
Endoscopic ear surgery (EES) is becoming popular in the last decade, which is identical with microscopic ear surgery (MES), However, EES showed some obvious advantages, such as a wider surgical view and less invasive incision from the external auditory canal [5][6][7], the advantages of EES seem to be especially remarkable in minimally invasive middle ear surgery, such as ossiculoplasty [6,8]. However, up to date, most TOD were treated by ossiculoplasty under microscope [1,2,9]. Moreover, TOD is usually accompanied with temporal bone fracture (TBF) and facial nerve (FN) injury in some patients [10,11], whether the exclusive ESS could be quali ed to manage these complex injuries is unknown. In order to explore the e cacy of EES for management of TOD and the accompanied injuries, we retrospectively collected and analyzed a series of TOD cases with intact TM which were performed EES for ossiculoplasty and FN decompression in a portion of cases accompanied with FN injury.

Clinical material
We retrospectively analyzed a series of patients who visited our hospital due to hearing loss after trauma between May 2017 and Jul 2019. The cases met the following criteria were included in this study: (1) a persistent conductive hearing loss after head or ear trauma; (2) an open external auditory canal and intact TM. Exclusion criteria: (1) TOD combined with otitis media. (2) Severe bony fracture of skull base, CSF leakage, or meningoencephalocele. (3) the surgery under microscope with a post-auricular incision or mastoidectomy was performed. (4) preoperative radiologic evidence suggested the FN injury involvement was not limited from the geniculate ganglion to pyramidal segment. (5) the case lost to follow-up and postoperative hearing outcome was not obtained. This study was approved by the Institutional Ethics Review Board of our institution and informed consent was obtained for our study from all participating patients.
A total of 85 ears were reviewed in this period, including 58 ears of traumatic TM perforation without TOD, 9 ears of traumatic TM perforation with TOD, 18 ears of TOD with intact TM. Finally, a total of 18 ears (16 cases) were included, the clinical characteristics and hearing status of these patients were summarized and analyzed ( Table 1). The patient population was composed of 12 males (13 ears) and 4 females (5 ears), with an age range of 7-54 years and a mean age of 36.6 ± 13.6 years. Fourteen patients showed unilateral TOD, and 2 patients showed bilateral TOD. The most common trauma etiology was fall down injury (9 cases), followed by tra c accident injury (5 cases) and blast injury (2 cases, Table 2). The average interval between trauma onset and surgery was 2.7 months (range: 0.5-24). Eight patients had complex injuries and were rstly treated in other departments, such as cerebral surgery, orthopedics department, maxillofacial department and so on.
Temporal bone HRCT scans and 3-D reconstruction of ossicular chain were routinely performed before surgery, TBF was observed in 13 ears, including 9 patients of longitudinal TBF, 3 patients of transverse TBF, and 1 patient of oblique TBF. Four patients (P3, P5, P13 and P15) presented with FN paralysis immediately after trauma onset, including 2 cases of House-Brackmann (H-B) grade (P5, P13), 1 case of H-B grade (P15), and 1 case of H-B grade (P3), all the cases were lack of response to corticosteroid therapy at least 10 days and radiologic evidence suggested FN injury. The preoperative pure-tone average (PTA) of four frequencies (0.5, 1.0, 2.0, and 4.0 kHz) ranged from 30 to 75 dB, the preoperative PTA and ABG were 48.1 ± 13.4 dB, and 32.4 ± 9.1 dB, respectively.

Surgical methods
All the EES procedures were conducted by the same senior surgeon (Z.C.) under general anesthesia via the transcanal approach with 0° and 30° endoscopes (2.7 mm of diameter and 11 cm of length) (Karl Storz, Germany). A posterior tympanomeatal ap was made and elevated to expose the tympanic cavity. The smaller amount scutum was removed using a curette for exposing the 3 ossicles, oval window and round window. The integrity and connection of the ossicular were explored. Various treatment strategies, including ossiculoplasty and FN decompression were applied depending on the intraoperative ndings: (1) a partial titanium ossicular replacement prosthesis (PORP) was used in the case of incus injury with normal stapes; (2) a total titanium ossicular replacement prosthesis (TORP) was used in the case of stapes suprastructure injury with a normal stapes footplate; (3) a titanium piston prosthesis was used in the case of stapes footplate injury with a normal incus; (4) FN decompression was performed in the case of FN paralysis induced by injury from the geniculate ganglion to pyramidal segment. A small piece of tragal cartilage was harvested and placed on the PORP/TORP prosthesis. The operation time and amount of bleeding were recorded. Two weeks after surgery, the external auditory canal was cleaned and the TM was examined with ear endoscopy in our outpatient department.
Interval hearing status was assessed by pure-tone audiometry. All patients were followed-up at least 6 months.

Statistical methods
A GraphPad Prism 7 statistical software was used for statistical analyses. A one-sample Kolmogorov-Smirnov test was used to analyze the normality of the data distribution. A paired-samples t-test or Wilcoxon signed rank test were used for comparisons before and after surgery. A P value < 0.05 was considered to indicate statistical signi cance.

Intraoperative ndings
A comprehensive description of the clinical characteristics, expletory ndings and treatment strategies were shown in Table 1 and Figure 1. It revealed that the incus injury was observed in 14 ears (77.8%), which was the most common type in this study, including 11 ears of incus dislocation involved both IMJ dislocation and ISJ dislocation, 2 ears of isolated ISJ dislocation, and 1 ear of the long process necrosis. Stapes suprastructure fracture was observed in 4 ears (22.2%), including 2 ears of stapes suprastructure fracture with normal footplate, 1 ear of suprastructure fracture with injured footplate, and 1 ear of suprastructure fracture accompanied with IMJ dislocation. All of the cases had edema mucus or brous tissue enveloped the injured ossicular. In the cases with FN paralysis, the injury sites were consistent with the suggestions of preoperative CT examination, including 3 ears of perigeniculate ganglion injury and 1 ear of tympanic segment injury. We also observed a case with partial missing lateral bony wall of external auditory canal (P7), which was constructed by a piece of tragus cartilage.

Surgical results of EES
The operation time ranged from 45 to 95 min, the average of operation time was 62.4 ± 15.5 min. The TM remained intact after surgery in all ears. An interval serial assessment of the hearing status was conducted postoperatively, a mean follow-up period was 18.4 months (range: 6.5-35.3). The mean PTA gain was 22.9 ± 9.5 dB, and the ABG was 22.2 ± 8.3 dB. The mean postoperative PTA was signi cantly reduced, and the mean postoperative ABG was obviously closed, respectively (P < 0.001) ( Table 3). The ABG closure to 20 dB or less and ABG closure to 10 dB or less were achieved in 18 ears (100%) and 14 ears (77.8%), respectively. The average bone conduction was not changed signi cantly (P = 0.236). (Table 3). The FN function recovered very well in all the 4 patients with FN paralysis, it showed 3 cases of H-B grade , and 1 case of H-B grade in 6 months after surgery.
According to the involved ossicles, the TOD cases were divided into two subgroups, incuse injury with normal stapes and stapes injury with or without incus injury, the preoperative PTA and ABG showed no statistical differences, respectively (P=0.126, P=0.386). Accordingly, the hearing reconstruction adopted PORP implantation for incus injury, and TORP implantation or stapedotomy and piston prosthesis implantation for stapes injury, respectively, no statistical differences were found in postoperative PTA and ABG, respectively (P=0.975, P=0.790) ( Table 4 and Fig 2).

Complications
The operative bleeding was less than 20 ml in all ears. A post-auricula approach, canalplasty or mastoidectomy was not necessary in any case. However, the chorda tympani nerve was transfected in 1 case (right ear of Case 6), in which the chorda tympani nerve was tightly adhered with the dislocated incus and the brous tissues surrounding it. No major complication, such as iatrogenic facial nerve paralysis, signi cant sensorineural hearing loss (≥ 15 dB) was observed.

Discussion
Ossiculoplasty was the main treatment for TOD, which was traditionally performed under microscope [1,2,9]. In recent years, EES for tympanoplasty and ossiculoplasty has increased annually, showing some obvious advantages, such as a wider and multiangle eld of view which allowed the surgeons to accurately assess and then reconstruct the ossicular chain [5][6][7][8]. However, ESS for the management of TOD have not been well studied, up to date, only Kim [12] reported 15 cases of TOD treated by ESS in 2019, the TOD cases with TM perforation were also included in their study, the diagnose of these cases was relatively easier, which could be observed directly using the endoscope through the perforation. As previously reported, the traumatic TM perforation of some TOD cases could be healed spontaneously in one or several weeks [13], which could lead to the diagnosis of TOD more di cult. To the best of our knowledge, this is the rst study focused on employing totally ESS for the management of TOD with an intact TM. Our research showed the ABG closure to 20 dB or less and ABG closure to 10 dB or less were achieved in 18 cases (100%) and 14 cases (77.8%), respectively, which was well in accordance with the study of Kim [12], and better than those of MES, Ghonim [2] reported 42 cases of TOD with intact TM, which were performed MES, the postoperative PTA and ABG were signi cantly improved to 20.2 ± 9.8 dB and 4.2 ± 7.3 dB, respectively. However, so far, there is absent of a study on direct comparison between ESS and MES for TOD, further studies employing a larger cohort and including MES are needed for validation. Moreover, we also analyzed the postoperative hearing outcomes between incus injury reconstrued by PORP implantation and stapes injury mainly reconstructed by TORP or piston implantation, no signi cant differences were found, the results were different from our previous study of ossicular chain malformation [8], the possible reason was that the piston reconstruction was also adopted for stapes injury with footplate anomaly (P4), as many studies reported, the piston prosthesis usually obtained a better hearing result [2,14,15].
In this study, head trauma through fall from a height or tra c accident was the most common cause of TOD, which was well in accordance with previous studies [2]. Incus is prone to be involved in injury as it was suspended in the tympanic cavity between the rmly anchored malleus and stapes [10], the injury types included isolated ISJ dislocation, isolated IMJ dislocation, and a complex of both. Isolated malleus handle and long process of incus fractures were rarely reported in previous reports [16][17][18], we observed a case of isolated long process necrosis (right ear of P16), and a case of stapes suprastructure fracture which was still suspended upon the stapes footplate (P14), both the ears was misdiagnosed by preoperative examination. HRCT is the preferred modality due to its ability to demonstrate the details of TOD and other associated injuries, such as TBF and FN injury. In this study, the diagnosis sensitivity of HRCT was 88.9% for TOD, which was in according with the previous reports [4], some TOD cases with the minor abnormalities were still missed in clinical practice. Therefore, exploratory tympanotomy is critical for the diagnosis, and could simultaneously perform ossiculoplasty to improve hearing.
TOD cases usually experienced a relatively long period of delayed diagnosis and treatment, it was still under controversy when to perform the tympanic expletory [19], suggestions of previous studies were from immediate exploration to 3 months, in this research the average interval between trauma onset and surgery was 2.7 months (range: 0.5 -24). Grant [19] reported conservative management was effective in traumatic TM perforation and hemotympanum, however, some TOD cases showed a worsen hearing, therefore, exploratory tympanotomy should be performed for the TOD cases with de nite radiologic evidences. Moreover, we found that the dislocated ossicles were usually enveloped with brous tissues, forming soft connection between the ossicles, which induced a relatively small average preoperative ABG of 32.4 ± 9.1 dB, compared with a larger average ABG of 63.2 ± 11.7 dB in the cases of isolated middle ear malformation in our precious study [8]. On the other hand, the brous tissue may induce the ventilation blockage and provoke a dysventilation of the middle ear [20]. Therefore, the surgery should be considered once HRCT scan suggested the disruption of ossicular with an ABG of 30 dB or more.
TOD was usually accompanied with TBF, perilymphatic stula, or FN injury [10]. The TOD cases accompanied with severe otologic complications should be treated timely, However, whether EES could be quali ed to manage these associated anomalies at the same time is yet unknown. To the best of our knowledge, this is the rst study to explore this question for TOD. According to previous reports, perilymphatic stula commonly occurred in the stapes and round window [21], though after carefully examination, no occurrence of perilymphatic stula was observed in our cases series, EES could provide a clear and magni ed view for checking these areas, and reconstruction if necessary [22], the characteristic of endoscopic view ensured the surgical procedures being minimally invasive. TBF was found in 13 ears in this research, 4 ears were accompanied with FN injury, the perigeniculate area and tympanic segment was most commonly involved in the injury, which was in according with previous researches [4,11]. Though the timing of treatment of traumatic FN paralysis was under controversy [23], the case of traumatic FN paralysis combined with TOD was a good indicator for timely surgery, especially for the cases with severe immediate-onset facial paralysis and lack of response to the corticosteroid therapy. A few approaches can be used for FN decompression depending on the injury sites, including transmastoid approach and the middle cranial fossa approach, however, it is a challenging to control these areas through traditional surgical approaches [11]. In this research, it showed EES was effective in the management of some selected FN injury, transcanal approach could allow the surgeons to control the whole tympanic segment, geniculate ganglion and lateral aspect of the labyrinthine portion, obviating the need of excessive removal of tympanic scute, canaloplasty, or a post-auricle incision and mastoidectomy, the FN function recovered very well in 6 months after surgeries in all cases. However, when radiologic evidence suggested the injury involved mastoid segment, the transmastoid approach should be conducted, 5 cases (excluded from this research) were performed transmastoid approach under microscope at the same period in our department, carefully preoperative imaging study was critical for selecting TOD patients with FN paralysis for EES exclusively or combined with transmastoid approach.

Conclusion
In conclusion, this research showed EES was effective in diagnosis and management of TOD with intact TM, and some accompanied otologic injuries. It showed favorable surgical results. EES provides an alternative method to manage TOD, with the advantage of excellent vision and less invasion, which could be promoted in clinical. Declaration Ethical approval and consent to participate This study was approved by the Institutional Ethics Review Board of the A liated Hospital of Zunyi Medical University and written informed consent was obtained for our study from all participating patients.

Consent for publication
All authors have agreed to publish this article.

Availability of data and materials
Data is available upon request by contacting the corresponding author.

Competing interests
The authors declare that they have no competing interests.  Tables   Table 1 Patients clinical characteristics, intraoperation ndings, surgery strategies and hearing outcomes PORP: partial titanium ossicular replacement prosthesis; TORP: total titanium ossicular replacement prosthesis; Incus dislocation involved both incudomalleolar joint and incudostapedial joint dislocation; ※ the cases accompanied with facial paralysis.
# the tragus cartilage was adopted to repair the EAC.   white arrow showed the necrosis of the incus long process, which was instead by the brous connection and appeared as normal; red arrow showed the minor missing of the incus long process after removal of the brous tissues. B: white arrow showed the disrupted incus which was enveloped by the edam mucus; red arrow showed the dislocated of incus which involved both the incudomalleolar joint and incudostapedial joint dislocation. C: red arrow showed the dislocation of incus, the FN nerve was depressed by the long process; white arrow showed the tissues connection between the lenticular process and the stapes. D: white arrow showed the normal connection of incudostapedial joint and the fractured stapes suprastructure which was suspended over the footplate; red arrow showed the complete fracture of stapes suprastructure after removal the dislocated incus.

Figure 2
Various strategies and skills of endoscopic ear surgery FN: facial nerve; MH: malleus handle; St: stapes; PORP: partial ossicular replacement prosthesis; A: white arrow showed the brous connection between the MH and incus long process; red arrow showed a small piece of cartilage placed on the PORP. B: white arrow showed the fracture of stapes suprastructure and the injured footplate; red arrow showed the piston implantation with the small fenestra stapedotomy. C: white arrow showed the injury of bone canal of tympanic segment FN; red arrow showed the local FN decompression. D: white arrow showed the FN injury in perigeniculate area, and the FN decompression from lateral aspect of the labyrinthine portion to the pyramidal segment; red arrow showed a piece of cartilage was used for reconstructing the fracture of skull base; yellow arrow showed a piece of perichondrium covered FN; blue arrow showed the ossiculoplasty using PORP implantation.