xRead - January 2023

Medical management of recurrent bouts of CSOM/ OE is of paramount importance in the early manage ment of PIMCF and may slow progression of disease; however, surgical therapy is the only therapeutic option that offers the potential to restore hearing if complete canal obliteration is present. Surgery may also be con sidered during the wet phase to halt progression of dis ease. 10 Outcomes of surgery in the management of PIMCF in the literature are generally reported to be favorable; however, the current body of evidence is limit ed by short-term follow-up, small cohorts, and retrospec tive study designs. Furthermore, few studies focus specifically on the postinflammatory subtype of MCF and instead combine data from patients with differing etiologies of MCF, which may introduce selection bias. Therefore, to begin to better understand the long-term efficacy of surgery, specifically for patients with the post inflammatory subtype of MCF, the objectives of the cur rent study were to: 1) review our short- and long-term hearing outcomes and rates of restenosis after surgical repair of PIMCF, and 2) systematically aggregate surgi cal outcomes data from available literature. This study was approved by the Medical University of South Carolina Institutional Review Board (Pro00046660). Patients who underwent surgical treatment for PIMCF between January 2000 and July 2015 were eligible for inclusion. Those diagnosed with MCF due to other causes (posttraumatic, postra diation, neoplastic, iatrogenic, idiopathic) were excluded. Addi tionally, all patients were required to have a minimum of one postoperative clinical follow-up with pre- and postoperative audiometric data available to be included in the study. Patient demographics, medical histories, and prior ear surgeries were recorded. Clinical characteristics, including the degree of canal stenosis and need for revision surgery, were documented. Audiometric Outcomes Postoperative audiometric data was collected at two time points. Short-term results were analyzed using the initial postop erative audiogram, which is routinely obtained within 6 months after surgery. Long-term hearing outcomes were assessed with each patient’s most recent audiogram obtained > 2 years after repair of PIMCF. Although there is significant heterogeneity in how authors define long-term results in the literature, our cutoff of 2 years is similar to other related reports. 8 Preoperative and postoperative audiometric data were recorded according to the Academy of Otolaryngology–Head and Neck Surgery standards. Specifically, air (AC) and bone conduc tion (BC) thresholds at 0.5, 1, 2, and 3 kHz were recorded and used to calculate four-tone pure-tone averages (PTA). If the 3 kHz threshold was not available, the average of the 2 kHz and 4 kHz thresholds was used. Air-bone gap (ABG) was calculated as AC PTA minus BC PTA. Change in ABG was calculated using preoperative ABG minus postoperative ABG. Preoperative and postoperative speech reception thresholds (SRT) were also recorded. Similarly, improvement in SRT was calculated as pre operative SRT minus postoperative SRT. The primary audiomet ric outcome was postoperative improvement in ABG, whereas secondary outcomes were postoperative SRT and postoperative achievement of ABG 10 dB and 20 dB. MATERIALS AND METHODS Patient Selection for Retrospective Review

Surgical Outcomes The primary surgical outcome was restenosis rate. To com pare short-term and long-term surgical outcomes, all patients who had clinical follow-up appointments within 2 years of surgery were included in the short-term analysis. Likewise, all patients who had data available from follow-up appointments > 2 years postop eratively were analyzed as part of the long-term cohort. Surgical Technique After induction of general orotracheal anesthesia, a split thickness skin graft (STSG) is harvested from the upper arm and placed in saline for use later in the procedure. The ear is injected with local anesthetic postauricularly and endaurally. A postauricular incision is made, and the ear is reflected in an avascular plane anteriorly. A self-retaining retractor is placed; periosteal incisions are made; and the soft tissue is elevated to the level of the ear canal. The ear canal is transected just later al to the fibrous/granulation tissue that is obliterating the EAC. The granulation and fibrous plug are removed from the EAC, working from lateral to medial. This is taken down to the level of the TM, and the soft tissue is dissected off the fibrous lay er of the TM. Ideally, the fibrous layer is left intact. If a perfora tion is created during dissection, tympanoplasty can be performed using a variety of standard techniques. A bony canalo plasty is routinely performed to remove all canal wall bulges or overhangs. The STSG is cut into two separate grafts; each graft is gent ly perforated to allow egress of any fluid collected beneath it. One graft is placed on the anterior canal wall and the other on the pos terior canal wall. Medially, both grafts overlay the TM, which is denuded of any residual (usually none) squamous epithelium. Laterally, the grafts are fashioned to overlap the native canal skin by 1 to 2 mm; no sutures are placed. Gelfoam is used to pack the EAC, and the postauricular incision is closed. Systematic Review and Meta-Analysis Two authors ( R . G . K . and A . A . O .) independently searched the PubMed, Scopus, and OVID/Medline databases following the Preferred Reporting Items for Systematic Reviews and Meta analyses statement (Fig. 1). Any discrepancies regarding inclu sion of any study were discussed among the authors to reach a mutual consensus. Studies meeting the following inclusion criteria were ulti mately selected: any randomized control, prospective cohort, and retrospective study describing any patient (pediatric or adult) receiving surgical intervention for PIMCF. Case reports, letters to the editor, abstracts, and book chapters were excluded. Articles describing canal stenosis alone without medial canal fibrotic plug or medical treatments of PIMCF were also excluded. No date range limitations were used. Data from the included articles were obtained indepen dently, including author, year of publication, number of patients and ears, patient demographics, audiometric data, and postop erative outcomes— including rates of restenosis. Data were col lected at two time points: short-term ( < 2 years) and long-term ( > 2 years). When individual patient data was reported, along with respective follow-up time for that given individual, this data was assigned to the appropriate short- or long-term group on an individual patient basis. In cases where individual follow up data was not reported, mean follow-up for the cohort was used to assign the entire study population into either the short or long-term groups.

Laryngoscope 127: February 2017

Keller et al.: Retrospective Review and Meta-Analysis 489