Can radiologic imaging replace second-look procedures for cholesteatoma?


  • Jerry W. Lin MD, PhD,

    1. The Bobby R. Alford Department of Otolaryngology–Head and Neck Surgery, Stanford University School of Medicine, Stanford, California, U.S.A
    Search for more papers by this author
  • John S. Oghalai MD

    Corresponding author
    1. Baylor College of Medicine, Houston, Texas, and The Department of Otolaryngology–Head and Neck Surgery, Stanford University School of Medicine, Stanford, California, U.S.A
    • The Department of Otolaryngology–Head and Neck Surgery, 801 Welch Road, Stanford University School of Medicine, Stanford, CA 94305-5739
    Search for more papers by this author

  • The authors have no funding, financial relationships, or conflicts of interest to disclose.


Cholesteatoma surgery is performed to eradicate disease, create a dry and safe ear, and restore hearing. The primary concern of the surgeon is to minimize the odds of recurrent and residual cholesteatoma. Recurrent cholesteatoma is a new cholesteatoma that forms from retraction of the tympanic membrane or ear canal skin. Recurrent cholesteatoma occurs at rates of 10% to 15% and are usually easy to diagnose in the office setting. Residual cholesteatoma forms from microscopic or gross disease left behind by the surgeon during the primary surgery. Reported rates of residual cholesteatoma vary from 5% to 63% in the literature. Residual cholesteatoma is not easily diagnosed in the office setting by otomicroscopic examination as the disease is often hidden within the mastoid or middle ear cleft.

Canal-wall-down procedures can reduce the odds of both recurrent and residual cholesteatoma but commit the patient to a lifetime of follow-up appointments and lifestyle restrictions.1 To circumvent this issue, many surgeons treat the chronic ear by means of a canal-wall-up procedure, done via a planned process of two surgeries, staged 6 to 18 months apart. The purposes of staging are to 1) allow mucosal healing to preserve the middle ear space, 2) reexamine the ear for residual cholesteatoma, and 3) restore hearing by means of ossicular chain reconstruction (OCR).

In severely diseased ears, when the mucosa is inflamed to such a degree as to threaten maintenance of a middle ear space and to obscure the certainty of complete cholesteatoma removal, the decision to stage the procedure is easy. It is presumed that a planned second-stage surgery will permit total resection of residual cholesteatoma pearls as well as achievement of a well-aerated middle ear cleft lined with healthy mucosa within which OCR can be expected to provide a better acoustic result (see dotted path in Fig. 1). In mildly diseased ears, the mucosa may appear generally healthy and the surgeon may have confidence in complete eradication of the cholesteatoma during the primary surgery. In these cases, it is an appealing option to save the patient from additional surgery by performing primary OCR with no definitive plan for a second look (see bold path in Fig. 1).

Figure 1.

Schematic for treatment plan of primary cholesteatoma. Dotted path represents conventional planned staged treatment. Bold path delineates a proposed treatment option for less severe disease.

Figure 2.

Utility of computed tomography (CT) for diagnosis of residual cholesteatoma. (a) Axial CT of a right temporal bone demonstrating an obvious focus of residual cholesteatoma in the mastoid cavity. (b) Intraoperative microphotograph of the residual cholesteatoma seen in (a). (c) Axial CT of a left temporal bone from a different patient demonstrating soft tissue within the mastoid cavity and surrounding the prosthesis in the middle ear. This material could represent cholesteatoma and/or scar tissue.


Because residual disease is difficult to detect in the office setting, a single-stage treatment plan relies on an effective means of evaluating the postoperative ear. Modern imaging technology may offer a less-invasive, reliable alternative to a second-look procedure in assessing for residual disease. In simple cases of recurrent cholesteatoma pearls in an otherwise well-aerated space, computed tomography (CT) imaging can be quite useful (see Fig. 2a and 2b). Limitations arise, however, regarding imaging resolution as well as inability to differentiate between recurrent cholesteatoma, inflamed mucosa, and scar tissue. As an example, Figure 2c is an equivocal case that was found to have scar tissue within the middle ear and mastoid during surgical exploration. In general, studies using CT imaging to diagnose residual cholesteatoma have been discouraging, with sensitivity and specificity rates around 50%.2

Newer magnetic resonance imaging (MRI) techniques may address the shortcomings of CT imaging and older MRI protocols. These techniques use diffusion-weighted fast spin echo (DW-FSE) sequences to detect cholesteatoma as a hyperintense lesion, in contrast to air, bone, granulation tissue and scar tissue, which are all hypointense. Dubrulle et al. compared imaging findings using such a protocol to intraoperative findings from second-stage (or third-stage) surgery in 24 patients (the reader is referred to this study for an excellent depiction of the MRI characteristics of cholesteatoma versus granulation tissue).3 Ten patients with negative imaging results were confirmed as disease-free during surgery. Residual cholesteatoma was correctly identified in 13 of 14 patients with positive imaging findings. The single false-positive resulted from the placement of bone dust in the middle ear to address a lateral semicircular canal fistula. Dhepnorrarat et al. used a similar imaging protocol to correctly identify seven of seven patients with residual cholesteatoma and 16 of 16 patients with no residual disease.4 The most notable shortcoming of these two studies was the lack of any patients with residual disease less than 5 mm in size. More recently, De Foer et al. used a similar protocol to evaluate patients with smaller lesions (nine of 10 smaller than 5 mm).5 Nine patients without intraoperative evidence of cholesteatoma were found to be disease-free by imaging. Of the 10 patients with residual cholesteatoma found during surgery, nine were correctly identified by DW-FSE MRI. The single missed lesion was a 2-mm cholesteatoma in a child with an imaging exam that was degraded by motion artifact. Of note, two adult patients with lesions as small as 2 mm were correctly identified by DW-FSE MRI. Combining the results of these three studies shows that DW-FSE MRI has sensitivity, specificity, positive predictive value, and negative predictive value scores of 97%, 97%, 97%, and 97%, respectively.


New MRI techniques that have improved sensitivity for residual cholesteatoma afford the surgeon more confidence when considering a single-stage surgical approach to the management of cholesteatoma. These studies, however, should not be considered conclusive because the sensitivity of MRI for a large number of patients over a long time period is not available. Given these shortcomings of the MRI experience, single-stage surgery with follow-up monitoring by imaging should presently be reserved for those patients found on initial surgery to have only mildly diseased ears from which cholesteatoma has likely been completely eradicated. Unfortunately, the evidence to date does not support more broadly applicable recommendations. Whether or not imaging is necessary or sufficient after a presumably successful procedure represents a knowledge gap that has significant clinical importance and requires a high-quality prospective randomized study to answer definitively.

Level of Evidence

The decision about whether or not to stage cholesteatoma surgery is often based on expert opinion, and very few studies address this question with good evidence. In this manuscript, the articles discussed included imaging studies providing level 2 evidence (development of diagnostic criteria on basis of consecutive patients with universally applied reference gold standard) and a review article providing level 5 evidence.