Customize your JAMA Network experience by selecting one or more topics from the list below.
Golz A, Goldenberg D, Netzer A, et al. Cholesteatomas Associated With Ventilation Tube Insertion. Arch Otolaryngol Head Neck Surg. 1999;125(7):754–757. doi:10.1001/archotol.125.7.754
To determine the incidence of cholesteatoma formation associated with ventilation tube (VT) placement and to identify and analyze the variables and risk factors that may predict or predispose to this complication.
We reviewed the medical records of 2829 children following VT insertion between the years 1978 and 1997 to obtain 1- to 20-year follow-up data.
Departments of Otolaryngology–Head and Neck Surgery and outpatient clinics of 2 tertiary referral academic medical centers.
A study population of 2829 children, ranging in age from 1.2 to 14 years (5575 ears), underwent a total of 6701 VT placements.
Main Outcome Measure
Cholesteatomas were considered a complication of VT placement whenever they developed at or near the site of the tube insertion.
Cholesteatomas directly attributed to VT placement occurred in 1.1% of the ears that were operated on. A higher incidence occurred (1) in children younger than 5 years, (2) when Goode T-tubes were used, (3) in cases with repeated insertions of tubes, (4) with intubation exceeding 12 months, and (5) in cases with frequent postoperative otorrhea.
Cholesteatoma formation associated with VT placement occurs in 1.1% of the ears that are operated on, and therefore it should be discussed with patients or parents prior to surgery. Periodic and long-term follow-up microscopic examinations of the eardrum should be performed in all patients following tubal extrusion or removal, especially in those at high risk for developing a secondary cholesteatoma, to detect this complication as early as possible.
VENTILATION tubes (VT) were reintroduced by Armstrong1 in 1954 for the treatment of chronic secretory otitis media. Since then, VT insertion has become the most effective surgical treatment for otitis media with effusion and the most commonly performed otologic procedure on children. Additional indications for VT placement are recurrent acute suppurative otitis media and atelectasis or retraction pockets of the tympanic membrane.2 Some authors have also suggested that VT placement may prevent the formation of cholesteatomas by ensuring long-term continuous middle ear ventilation.3 With the widespread use of VTs, the undesirable effects of these tubes are becoming japparent. As with any medical or surgical intervention, there are occasional complications due either to the operative procedure itself or to the anatomical or physiological effects of the tubes. Hearing loss, persisting perforations, tympanosclerosis, atrophy, atelectasis, and the development of cholesteatomas have all been attributed to the use of VTs.4-6 In recent years there has been much controversy over the widespread use of VTs, with some authors describing the procedure as having reached "epidemic proportions" and others decrying the procedure as an "assault on the eardrum."5
The development of a secondary cholesteatoma at the site of the tube placement is a very uncommon complication, ranging in most series from 0% to 1%.3,7-10 The average incidence reported in a survey of 12 studies was 0.5%,11 although a few studies reported a higher incidence that ranged from 1.6% to 7%.12-16 In a survey of 312 otologists who used VTs, Armstrong17 noted that 25 of these physicians reported a total of 39 instances of middle ear cholesteatomas (8%), but few attributed the development of the cholesteatoma directly to the use of the tube.
Isolated case reports have demonstrated a direct relationship between the use of a VT and the later development of a cholesteatoma. Hughes et al7 found 2 cholesteatomas that they believed to be related to tube placement. McLelland,8 in a study of 697 ears, reported 1 case of a cholesteatoma that occurred secondary to the tube insertion process and presented immediately beneath the drum surface where the tube had been placed. Fernandez-Blasini18 reported a case in which a "pearly" cholesteatoma cyst developed in the area of insertion of a VT. Schwartz and Linde19 and Glasgold20 both presented an iatrogenic implantation cholesteatoma as an unusual complication of tympanostomy tubes. Schwerdtfeger21 described 12 cases of cholesteatomas related to tympanostomy tubes that were a result of either improper myringotomy or poorly controlled removal of the tube. Pahor22 presented a case of an intratympanic keratoma following grommet insertion. Herdman and Wright23 also found an anterior tympanic cholesteatoma behind an intact eardrum at the site of previous VT placement in a child, which they called a "grommet cholesteatoma." On the other hand, Padgham et al24 and Kinsella25 demonstrated that there has been no change in the incidence of cholesteatoma formation in their patients despite a considerable increase in the use of VTs. Kokko and Palva9 have also stated that, to a critical observer, cholesteatomas occurred in most cases despite tympanostomy tubes, not because of them.
There is also some controversy regarding the development of a cholesteatoma following middle ear intubation, although in theory, the conditions for the development of a secondary cholesteatoma are perfect both when a VT is in place and when a perforation or atrophic scar remains at the intubation site. A cholesteatoma may arise (1) as a direct complication of VT placement in a retraction pocket in an atrophic and flaccid area in the drum that progresses to the point of debris accumulation, (2) as a result of shedding and implantation of epithelial cells into the middle ear, or (3) due to ingrowth of squamous epithelium from the perforation margin to the undersurface of the drum, which is facilitated by the flanges of the tube.7,9-11,23,26 Mortensen and Lildholdt27 concluded that a complicated interaction of various processes may result in the development of a cholesteatoma at the tube site and that the presence of the tube may have a contributory effect.
The purpose of this retrospective study was to determine the actual incidence of cholesteatoma formation directly attributable to VT placement. The study includes a large series of patients over a long period of follow-up to better to identify and analyze the variables and risk factors that may predict or predispose to this complication.
The study consists of a retrospective analysis of myringotomies with VT insertions performed in 2 academic medical centers over a 20-year period from 1978 through 1997. Preoperative evaluations, operative reports and records, and follow-up evaluations were studied.
Review of the notes yielded the following specific data: (1) sex; (2) age at the time of surgery; (3) indications for surgery; (4) previous tube placement; (5) type of tube selected; (6) frequency of postoperative otorrhea; (7) duration of ventilation; (8) whether spontaneous extrusion occurred or intentional removal of the tube was performed; (9) time span of postoperative follow-up; and (10) the presence of a cholesteatoma and its site in relation to prior VT placement. Statistical analysis of the incidence of post-VT cholesteatoma formation was evaluated by the χ2 test and the Student t test, with P≤.05 regarded as significant.
A total of 4247 children underwent myringotomy and VT insertion during the years 1978 through 1997. We were able to observe 3362 of these children for a period of 1 to 20 years. Children in whom the indication for tube placement was a retraction pocket or atelectasis of the drum were excluded from the study. We also excluded ears in which the tubes were still in place and ears in which an attic cholesteatoma had developed.
The remaining 2829 children (5575 ears) underwent a total of 6701 tube insertions. There were 1538 boys and 1291 girls (1.19:1 ratio). The age at the time of surgery was 1.2 to 14 years (mean±SD age, 4.8 ±2.6 years). In 2746 children, tubes were inserted in both ears, while 83 children had their tubes placed only in 1 ear. We used 6269 "homemade" polyethylene tubes in 5143 ears, and 432 Goode T-tubes were inserted in 432 ears. These tubes were inserted primarily in 168 ears, and as an alternative to the polyethylene tubes in 264 ears following the polyethelene tubes' failure to maintain a prolonged middle ear ventilation. In 4793 ears, tubes were inserted only once; in 567 ears, tubes were placed twice; and 215 ears had 3 or more tube insertions (range, 3-6; average, 3.6). It is therefore more convenient to express the results in terms of ears rather than patients.
The indications for tube placement were chronic, persistent middle ear effusion unresponsive to medical therapy in 4937 ears and recurrent acute suppurative otitis media in 638 ears. On average, the polyethylene tubes were in place for 8.8 months (range, 1-26 months), while the Goode tubes maintained their position in the middle ear for an average period of 20.4 months (range, 8-50 months). The time the tube was in place was defined as the time between the date of insertion and the date the tube was last seen by a physician to be in place.
Cholesteatomas were considered a complication of VT placement when they developed behind an intact drum or next to a perforation at or near the site of the tube insertion, in the mesotympanum or hypotympanum. The diagnosis of a cholesteatoma was made according to the microscopic findings on examination of the drum by computed tomography of the ears and was confirmed in all of the patients during surgery on the affected ears.
Sixty-two such cholesteatomas (1.1%) were diagnosed during the follow-up period, 33 cases among the boys and 29 among the girls, with no sex predilection. Most of the cholesteatomas, 48 cases, were found in children who were operated on before age 5 years, as compared with only 14 cases that developed in children who were operated on after this age. This difference is statistically significant (P<.001).
Forty-six cholesteatomas (74.2%) were found in ears that were operated on during the years 1978 through 1987, while only 16 (25.8%) were diagnosed in those ears in which tubes were placed during the years 1988 through 1997. The number of ears operated on was almost equal in these 2 periods (2728 and 2857, respectively). The significance of the different rates is unknown, but since the operating techniques have not changed over the years, we assume that a longer follow-up will reveal more cholesteatomas in those ears that were operated on between 1988 and 1997.
Forty-nine cases of cholesteatoma were found in the ears treated with the polyethylene tubes (0.95%) and 13 in those ears in which the Goode T-tubes were used (3%). This different incidence is statistically significant (P<.05).
The incidence of posttympanostomy cholesteatoma formation was equal in ears in which the indication for tube placement was persistent middle ear effusion and in those ears in which the indication was recurrent acute otitis media. In cases in which no episode or up to 2 episodes of postoperative otorrhea occurred, the incidence of cholesteatoma formation was 1.08% (57 of 5253 ears). But when there were 3 or more such episodes, a cholesteatoma was found in 1.55% (5/322) of the ears. Although this difference is not statistically significant, it shows that a cholesteatoma will tend to develop in an ear with recurrent episodes of otorrhea rather than in a "dry" ear.
Regarding the number of previous tube insertions, the incidence of cholesteatoma formation was higher in ears that were operated on 3 or more times (6/215 [2.8%]), as compared with ears that had only 1 or 2 tube placements (56/5360 [1.04%]). This difference is statistically significant (P<.01).
In 52 cases (1.4% of 3724 ears) of cholesteatoma, the tubes remained in position for 1 year or more, as compared with only 10 cases (of 1851 ears [0.54%]) that developed in ears in which the tubes remained less than 1 year. This difference is also statistically significant (P<.01). The reason for this difference is related not only to the prolonged intubation period but also to the different types of tubes used; "short-term" vs "long-term" tubes.
There was no difference in the incidence of postoperative cholesteatoma formation between those ears from which the tubes extruded spontaneously and those ears from which the tubes were removed. Over the 20 years of this study, we performed more than 1000 myringotomies without tube placement in 454 children ranging in age from 6 months to 7.5 years. A cholesteatoma at or near the site of the myringotomy was not encountered in any of these cases.
Since Armstrong1 reintroduced the use of VT in 1954, myringotomy with VT placement has become the most commonly used surgical treatment for persistent middle ear effusion, for recurrent acute otitis media, and in cases of adhesive otitis media to prevent cholesteatoma formation. Although considered a simple procedure with substantial benefits, VT insertion may cause complications and recently has been the subject of much contention.5
This study focuses on an uncommon but important potential complication of VT insertion—the formation of a secondary cholesteatoma at the site of tube placement. Although most authors did not find this complication in their studies, we, as a result of a large cohort with 20-year follow-up, were able to demonstrate that VT insertion, rather than preventing cholesteatomas, may actually cause their formation in up to 1.1% of the ears operated on. Use of the procedure thus merits serious consideration.
This study examined different variables that may influence the incidence of post-VT cholesteatoma formation. We demonstrated that this type of cholesteatoma has a higher incidence under certain circumstances: (1) when the tubes are inserted in children younger than 5 years; (2) with repeated tube placements; (3) with the use of "long-term" tubes; (4) whenever the tubes are in place for prolonged periods; and (5) in cases of frequent otorrhea. Herdman and Wright23 noted that children who require multiple tube insertions constitute a high-risk group for cholesteatoma development, and Inglis26 suggested that cholesteatoma formation may depend on the type of the tube used or the length of time it is in place, with longer retention associated with higher rates of cholesteatoma formation.
The exact mechanism by which a cholesteatoma forms at the site of tube placement is unknown. Because such cholesteatomas were not encountered following myringotomy alone, we cannot accept the seeding or implantation theories. We assume that a cholesteatoma may develop either in a retraction pocket in an atrophic scar or due to reverse epithelial ingrowth from the perforation margins to the undersurface of the drum, directed and enhanced by the tube's flanges. The fact that post-VT cholesteatomas were found more frequently in ears treated with Goode T-tubes and in ears intubated for prolonged periods supports this latter theory.
Since the actual incidence of cholesteatoma formation associated with VT placement is as high as 1.1%, complacency regarding the possibility of this complication should be condemned, and the risk should be discussed with patients or parents prior to surgery. Periodic and long-term follow-up microscopic examinations of the eardrum should be performed in all patients following tubal extrusion or removal, especially in those at high risk for developing a secondary cholesteatoma, to detect this complication as early as possible.
Accepted for publication February 19, 1999.
Reprints: Avishay Golz, MD, Rambam Medical Center, Department of Otolaryngology–Head and Neck Surgery, PO Box 9602, Haifa 31096, Israel (e-mail: firstname.lastname@example.org).