- Email: [email protected]
Mid-temporal pericranial and inferiorly based periosteal flaps in mastoid obliteration
Otolaryngology–Head and Neck Surgery (2007) 137, 906-912
ORIGINAL RESEARCH—OTOLOGY
Mid-temporal pericranial and inferiorly based periosteal flaps in mastoid obliteration Matthew Yung, PhD, FRCS, and Phillip Smith, PhD, Ipswich, UK OBJECTIVE: To study the outcome of mastoid obliterations where mid-temporal pericranial flap and inferiorly based periosteal flap were used. STUDY DESIGN: A prospective longitudinal study with a minimum follow-up of 1 year. SUBJECTS AND METHODS: Between 2000 and 2005, 102 mastoid obliterations were performed on 96 patients, the majority having hydroxyapatite granules as fillers. Of these, 31 were primary obliterations and 71 were secondary obliterations. Thirty were on children under 16. RESULTS: Using Kaplan-Meier analysis, the percentage of ears that were free of residual cholesteatoma at 5 years was 91% (95% CI: 83.1%, 100.0%). There was no recurrent cholesteatoma. Cutoff analysis was used to study other parameters at 5 years. The 5-year otorrhea rate was zero, and 96% of the cavities became completely epithelialized. Twenty out of 24 ears were water resistant, and 50% had air-bone gaps of 20 dB or better. CONCLUSION: The outcome of this technique compares favorably to other reports. © 2007 American Academy of Otolaryngology–Head and Neck Surgery Foundation. All rights reserved.
M
astoid obliteration is an established technique for the treatment of chronic infection of the mastoid bone.1 A variety of techniques of mastoid obliteration have been described. These are all based on the same principle: the cavity is reduced using either biological tissues or biocompatible materials as fillers and a soft tissue flap is then used to cover the obliteration.2 For some of these techniques, the filler and the covering material is a single soft tissue flap.3,4 The obliterated cavity must be lined by keratinizing squamous epithelium for it to remain dry and water resistant. The migration of squamous epithelium over the soft tissue flap begins at the remnant of pre-existing meatal skin or tympanic membrane. It could be impeded by a number of unfavorable factors, such as localized infection, ischemia, or necrosis of the flap. Theoretically, a healthy vascularized flap should re-epithelialize faster than an ischemic flap. In general, soft tissue flaps used in mastoid obliteration can be categorized into pedicled flaps (eg, superiorly based musculo-periosteal flap,4 anteriorly based musculo-perios-
teal flap3) or axial flaps (eg, mid-temporal pericranial flap,2 temporoparietal fascial flap5). The senior author has been using a combination of two flaps, an inferiorly based periosteal flap (pedicle flap) and a mid-temporal pericranial flap (axial flap), to line the obliterated cavities since 2000. The present report is a prospective longitudinal study of the ears following mastoid obliteration.
PATIENTS AND METHODS Ear Audit Clinic The senior author has had a policy of a yearly review of patients following mastoid surgery. A weekly ear audit clinic was set up and the senior author personally evaluated the patients at each postoperative follow-up. The clinical information was recorded on a data proforma.
Operative Technique for Mastoid Obliteration A postaural incision was used in all patients to allow exposure and harvesting of the inferiorly based periosteal flap and the mid-temporal flaps. The postaural skin flap was lifted forward to expose the periosteum over the mastoid bone. An inferiorly based periosteal flap was then created with a broad base near the tip of the mastoid process. In most cases, the senior author incorporated the adjacent temporalis fascia into the apex of the periosteal flap to make it longer (Fig 1A). The mid-temporal pericranial flap is based on the midtemporal artery. The flap has been previously described in detail by Black.2 The temporalis muscle overlying the pericranium was dissected off to expose the pericranium. The middle temporal artery has a dual origin from both the deep temporal and the superficial temporal arteries. Besides the main axial vessels, a network of smaller vessels is also present. The middle temporal vessels are usually readily identifiable as a leash of vessels running posterosuperiorly. The flap was fashioned according to the path of the middle temporal vessels and was made as long as possible and at least 40 to 50 mm wide.
Received June 24, 2007; revised August 20, 2007; accepted September 9, 2007.
0194-5998/$32.00 © 2007 American Academy of Otolaryngology–Head and Neck Surgery Foundation. All rights reserved. doi:10.1016/j.otohns.2007.09.014
Yung and Smith
Mid-temporal pericranial and inferiorly based . . .
907
Figure 1 (A) The outline of the inferiorly based periosteal flap. A small piece of temporalis fascia is incorporated into the tip of the flap. zg, groove across the root of the zygoma; smc, supramastoid crest; pf, periosteal flap. (B) The outline of the mid-temporal flap. The inferiorly based periosteal flap is already prepared. zg, groove across the root of the zygoma; mtf, mid-temporal flap; pf, periosteal flap. (C) The mastoid cavity is obliterated with hydroxyapatite granules and cartilage plates. The inferiorly based periosteal flap and the mid-temporal flaps are used to cover the fillers in the mastoid cavity. mtf, mid-temporal flap; pf, periosteal flap.
In most cases, the senior author enlarged the flap by extending it posteriorly and incorporating part of the pericranium overlying the occipital bone into the flap (Fig 1B). This resulted in a big flap that was long enough to reach the annulus and cover the whole attic area. A revision mastoidectomy was then performed to remove disease in the mastoid bowl and eradicate any infected mastoid cells. Healthy-looking epithelial lining of the mastoid cavity was elevated from the underlying bone and used
to reline the ear canal. The three semicircular canals, dura, and the sigmoid sinus were skeletonized to facilitate exenteration of the diseased mastoid cells. Porous hydroxyapatite granules (0.5-1.0 mm diameter; Pentax, Smith & Nephew Richards, Tennesee, USA) were soaked in antibiotic solution before they were put into the mastoid cavity. The bulk of the granules were placed within the mastoid cavity, with only a thin layer put into the epitympanic space.
908
Otolaryngology–Head and Neck Surgery, Vol 137, No 6, December 2007
A large piece of conchal cartilage was harvested and sliced into thin plates using the Kurz Precise cartilage knife (Kurz, Dusslingen, Germany). The cartilage plates were used to tile over the hydroxyapatite granules, starting at the epitympanum. Thicker sheets were used to reconstruct the attic wall, supported behind by the hydroxyapatite granules. Once all the cartilaginous “tiles” were in place, the inferior portion of the reconstructed area was covered by the inferiorly based periosteal flap. The main part of the obliterated area was covered by the mid-temporal pericranial flap, from the epitympanum to the floor of the new canal. Part of the mid-temporal pericranial flap inevitably would lie over the inferiorly based periosteal flap (Fig 1C). Any existing healthy-looking mastoid lining was placed on top of the mid-temporal pericranial flap. For most revision mastoidectomies, a meatoplasty would already be present. Otherwise, a meatoplasty was performed to provide a funnel-shaped opening to the canal.
Cohort of Obliterated Cavities Between 2000 and 2005, a total of 102 mastoid obliterations for chronic otitis media were performed on 96 patients (age range 5-80 years, 64 males and 32 females). A total of 30 obliterations were performed on children under age 16 and six patients had bilateral operations. The minimal follow-up period for each ear was 12 months. Of these 102 mastoid obliterations, 71 were on discharging mastoid cavities (secondary obliteration) and 31 were primary reconstructions of the cavity following a canal wall down mastoidectomy (primary obliteration). Ninety-five cavities had hydroxyapatite granules and cartilage plates as fillers; five smaller cavities only had cartilage as fillers; and two ears had the soft tissue flaps placed directly into the mastoid cavity with no fillers. The patients were followed up at 3 weeks, 6 weeks, 3 months, and 6 months after the operation and then on a yearly basis. At follow-up visits, a microscopic examination of the ear and pure tone audiogram was performed. Patients were encouraged to go swimming to test water resistance. Any otological symptoms from swimming, hair washing, or shower were documented. For ears that were obliterated by hydroxyapatite granules, an interval CT scan of the mastoid bone was routinely performed at 12 months.
Audit Information collected for the present study included the time interval for the ear to become epithelialized, incidence of postoperative otorrhea, patient perception of water-resistant properties, residual or recurrent cholesteatoma, and hearing results. Residual cholesteatoma usually took the form of an epithelial pearl due to remnants of the cholesteatoma matrix left behind the tympanic membrane or within the mastoid bone. The interval CT scan results at 12 months and any subsequent ear operations were also recorded. The residual cholesteatoma rates were studied using KaplanMeier analysis (MedCalc for Windows). Cut-off analysis
was used to study the other surgical outcomes at 5 years, including hearing outcomes. The outcome of primary obliteration and secondary obliteration is compared. At the outset of this study, an audit approach was deemed to be most appropriate. Consequently Institutional Review Board or UK equivalent of approval was not sought. The conduct of this audit was consistent with UK best practice.
RESULTS With the benefit of a dedicated ear audit clinic, 84 of the 96 patients (90 of the 102 ears) attended all the scheduled follow-up visits. A total of 101 ears had been followed up for at least 1 year, 73 ears for 2 years, 48 ears for 3 years, 33 ears for 4 years, and 24 ears for 5 years or longer. Of the 102 ears, 90 had an interval CT scan at 12 months. The purpose of the CT scan was to identify areas of low CT density (suggestive of residual cholesteatoma pearls) in a background of high CT density (hydroxyapatite granules).6 Six ears did not have CT scanning because hydroxyapatite granules were not used in their obliterations. It was felt that the differences in CT densities between residual cholesteatoma pearls and surrounding soft tissues did not give sufficient contrast for little pearls to show up on the scan. For the other six ears, five missed the scan and one patient had left the country. There was no recurrent cholesteatoma in the present cohort, but an attic cavity redeveloped in one ear due to loss of the hydroxyapatite granules in the epitympanum. The attic region was successfully reobliterated. Residual cholesteatomas were detected in five ears; three were in the middle ear (detected at 24, 40, and 51 months postobliteration) while the other two residual cholesteatomas were in the epitympanum (detected 32 and 66 months postobliteration). All the residual cholesteatomas were detected during routine follow-up and appeared as pearls behind the tympanic membrane or the cartilaginous attic wall. None of the residual cholesteatomas were picked up by the interval CT scans at 12 months. Of the 90 interval CT scans (12 months postobliteration), only one showed a shadow beneath the hydroxyapatite granules. This ear was re-explored, and the shadow was found to be due to fibrous tissue rather than cholesteatoma. Using Kaplan-Meier analysis, the percentages of ears within the present cohort that were free of residual cholesteatoma at 3 years and 5 years were 97.0% (95% CI: 92.9%, 100.0%) and 91% (95% CI: 83.1%, 100.0%), respectively. Table 1 depicts the outcome parameters of the ears in the present cohort. The outcome of primary obliterations is compared to that of secondary obliterations. Overall, 98 of the 102 ears (96%) had complete re-epithelialization of the cavity, and over 80% of the ears achieved that within 12 weeks following obliteration. A total of five ears (4.9%) suffered from intermittent otorrhea following obliterations including two incidences
Yung and Smith
Mid-temporal pericranial and inferiorly based . . .
909
Table 1 Outcome of using combined mid-temporal and inferiorly based periosteal flaps as soft tissue covering in mastoid obliteration Primary obliteration (n⫽31)
Parameters
Secondary obliteration (n⫽71)
Total cohort (n⫽102)
5-y cut-off analysis (n⫽24)
Children : adults Time to become fully epithelialized (weeks) Obliterate using HA granules Follow-up periods (months)
17 : 14 Median: 8 wks; 30/31 within 12 wks (96.8%) 27 ears
13 : 58 31 : 71 7 : 17 Median: 8 wks; 56/71 Median: 8 wks; 86/102 Median: 8 wks; 23/ 24 within 12 wks within 12 wks within 12 wks (95.8%) (84.3%) (78.9%) 68 ears 95 ears 22 ears
Median: 32 mos; range: 12-75 mos
Median: 34 mos; range: 12-80 mos
Median: 67 mos; range: 60-80 mos
12-month interval CT scan Residual cholesteatoma Recurrent cholesteatoma Postoperative otorrhea
27/31 completed
Median: 35 mos; range: 12-80 mos (1 immigrated at 8 mos) 63/71 completed
90/102 completed
21/24 completed
0 children, 2 adults
2 children, 1 adult
2 children, 3 adults
2 children, 2 adults
0
0
0
0
0
Single episode: 1; intermittent: 5 (4.9%)
0
Appearance of cavity
30 complete epith. (96.8%); 1 incomplete epith. 27 swim/hair wash (87.1%); 2 unknown
Single episode: 1; intermittent: 5 (6.9%) (2 myringitis; 1 retraction pocket; 1 meatal stenosis; 1 incomplete epith.) 68 complete epith. (95.8%); 3 incomplete epith. 51 swim/hair wash ⫹ 4 hair wash (77.5%); 11 unknown 2 incomplete epith. ⫹ 1 meatal narrowing ⫹ 2 myringitis (7.0%)
Water resistance
Not water resistant
1 incomplete epith. ⫹ 1 TM perforation (6.5%)
Subsequent ear operation
3 second-look ⫹ 1 myringoplasty ⫹ 2 removal of residual cholest. (19.4%)
2 second-look ⫹ 1 myringoplasty ⫹ 3 removal of residual cholest. ⫹ 1 repair retraction pocket ⫹ 2 meatoplasties ⫹ 1 ossiculoplasty ⫹ 1 skin grafting of cavity (15.5%)
98 complete epith. (96.1%); 4 incomplete epith. 82 water-resistant (80.0%); 13 unknown
23 complete epith. (95.8%); 1 incomplete epith. 19 swim/hair wash ⫹ 1 hair wash (83.3%); 1 unknown 7 non-water-resistant 1 incomplete epith. (6.9%) ⫹ 1 TM perforation ⫹ 1 myringitis (12.5%) 3 second-look ⫹ 2 5 second-look ⫹ 2 myringoplasties myringoplasties ⫹ 5 ⫹ 4 removal of removal of residual residual cholest. cholest. ⫹ 1 repair ⫹1 retraction pocket ⫹ ossiculoplasty ⫹ 2 meatoplasties ⫹ 1 1 skin graft of ossiculoplasty ⫹ 1 cavity (45.8%) skin grafting of cavity (16.7%)
HA, hydroxyapatite; TM, tympanic membrane; epith., epithelialization; cholest, cholesteatoma; mos, months; wks, weeks.
of myringitis and one incidence for each of the following: attic retraction pocket, meatal stenosis, and incomplete epithelialization of the obliterated cavity. Most patients tried swimming or washing their hair without ear protection to test water resistance; 81 ears were
water resistant while 7 were not. The reasons for the non– water resistance are given in Table 1. The water-resistant property was unknown in 13 ears because of coexisting pathology in the opposite ear, or patients’ reluctance to change the old habit of wearing ear plugs.
910
Otolaryngology–Head and Neck Surgery, Vol 137, No 6, December 2007
A total of 17 ears had further surgery during the observation period for a variety of reasons. They were secondlook tympanoplasties (five ears), surgery to remove residual cholesteatoma pearls (five ears), myringoplasties (two ears), meatoplasties (two ears), ossiculoplasty (one ear), split skin grafting of nonepithelialized areas in the cavity (one ear), and repair of attic retraction pocket (one ear). None of the patients had major postoperative complications, such as dead ear, facial palsy, or prolonged dizziness. Trimming of the soft tissue flap was performed on three ears because of tip necrosis. It is recognized that outcome reports on cholesteatoma surgery should be based on long-term studies. Therefore a cut-off analysis on the subset of ears that had a minimum follow-up period of 5 years is included in Table 1. The subset comprised 24 ears, with a median follow-up period of 67 months. The dry ear rate was 100%, with 96% of the obliterated cavities completely re-epithelialized. Also, 83% of the ears in this subset were water resistant. The presentation of the hearing results is less straightforward. The authors decided to present the 5-year hearing results to show what has been achieved, accepting that 11 out of 24 ears had more than one operation on the ear (Table 2). The average 5-year air-bone gap was 22.29 dB (95% CI: ⫺2.33 dB, 46.92 dB). Two patients had worsening of the air-bone gaps. Overall, 5-year air-bone gaps of 20 dB or better were achieved in 50% of the ears.
DISCUSSION A number of soft tissue flaps have been described for mastoid obliteration.2 The choice is often down to personal preference, rather than based on clinical evidence. When a soft tissue flap is placed over living bone, it gets nourishment directly from the bone. Indeed, free temporalis fascia graft is commonly used by surgeons to line the mastoid cavity.7 When a soft tissue flap is placed over nonvascular materials, such as bone paste, hydroxyapatite, or titanium canal prosthesis, it needs a pedicle or feeding vessels for nourishment. When free grafts are placed over nonvascular materials, they could break down and lead to exposure of the underlying materials or otorrhea.8-10
Table 2 Five-year hearing results following mastoid cavity obliteration (nⴝ24) Number of ears (%) Air-bone gap
Preoperative
0-10 dB 11-20 dB 21-30 dB ⬎30 dB
1/24 4/24 8/24 11/24
(4.2%) (16.7%) (33.3%) (45.8%)
Postoperative 5/24 7/24 5/24 7/24
(20.8%) (29.2%) (20.8%) (29.2%)
The senior author has been using hydroxyapatite granules as fillers in mastoid obliteration since 1990. Initially, the granules were covered by just an inferiorly based periosteal flap. It was found that the flap was not long enough to reach the annulus and the attic area. The senior author modified his techniques since 2000. Cartilage slices were used to cover the hydroxyapatite granules to smooth out the crevices and reinforce the attic reconstruction. In addition to the inferior-based periosteal flap, a superiorly based midtemporal pericranial flap was also used to provide vascular covering of the cavity. The present report is an observational study on 102 ears using this technique. The senior author favors the mid-temporal pericranial flap because of its good blood supply. It is, however, thin and can only be used as a covering flap rather than as filler. The reason for using two flaps is to provide extra cover over the granules. It is not yet clear how dependent the pericranial flap is on the middle temporal artery. Ramsey et al11 used an inferiorly based periosteal-pericranial flap to cover bone paste in mastoid obliteration and still reported a 65% complete dry ear rate. That indicates that the nourishment of the pericranium is not totally dependent on the middle temporal vessels. A concern with mastoid obliteration is that residual cholesteatomas may be buried underneath the fillers. Secondlook operation12 has been suggested as a possible way to detect these residuals. However, it has been reported that many ears developed residual cholesteatoma even when they have been cleared at the second-look operation.13 In the present study, the residual cholesteatoma only became apparent several years after the obliteration. None were detected by interval scanning at 12 months. It may well be that scanning of the obliterated ears at 12 months is too early. The authors are currently considering having the interval scan scheduled at 24 months instead. There is emerging evidence that MRI scan can differentiate cholesteatoma from other soft tissues within the mastoid bone but is still not specific enough to replace a second-look operation.14 Whatever policy one uses to detect residual cholesteatoma, there is an almost universal agreement that long-term follow-up of these patients is necessary.15,16 The aim of mastoid obliteration is to stop otorrhea and render the ear water resistant. In the present cohort, 95% of the ears were free of otorrhea, and 96% of the cavities were completely re-epithelialized. For the ears with otorrhea resulting from myringitis or incomplete epithelialization, they were treated with 2% acetic acid eardrops or sprays, and topical antiobiotic-steroid eardrops only if necessary. For the patients who had persistent otorrhea, revision surgeries were offered, such as skin grafting of the nonepithelialized areas. In the present cohort, over 80% of the ears were completely re-epithelialized within 12 weeks following the obliteration. Eighty-two out of 89 ears were found to be water resistant, and most patients could enjoy water sports. In order to facilitate the comparison of our results to other reports, a literature review was performed. A Medline search
Yung and Smith
Mid-temporal pericranial and inferiorly based . . .
911
Table 3 Literature review
Author
Size of cohort
Follow-up periods
Fillers
Covering tissue
Black18
124
Average 3 years
HA canal prosthesis
Mid-temporal flap
Roberson8
62
0.2-54.8 months
Bone paste
D’Arc9
72
1-158 months; average 46 months
Biphasic calcium phosphate granules
Free temporalis fascia Free temporalis fascia
Estrem10
31
1-74 months
HA canal prosthesis; HA granules
Free homograft dura ⫹ temporalis fascia
Leatherman16
11
6-20 months
Demineralized bone matrix ⫹ cartilage
Nyrop19
27
10-13 years
Palva20
185
1-11 years; average 5 years
Bone chip and paste Bone chip and paste
Free perichondrium graft ⫹ Palva flap Muscle flap
Ramsey11
60
12-80 months; average 32 months
Bone paste
Inferiorly based periostealpericranial flap ⫹ split skin
Present series
102
12-80 months; median 34 months
HA granules ⫹ cartilage
Mid-temporal flap ⫹ inferiorly based periosteal flap
Palva flap
Outcome Implant exposure 2.5%; avascular necrosis 3%; infection 4%; myringitis 8% 8/62 still discharging (13.8%) 19% still otorrhea at 12 months (a number of patients had leakage of granules) HA canal wall 25% failure; HA granules 13% failure (failure include extrusion of HA, and granulation tissue on the lining) 2/11 prolonged granulation, thought to be due to exposed bone matrix 89% dry ear; average no. of op ⫽ 1.2 10% had revision surgery; 80% remained dry 10% persisent discharge; 25% intermittent discharge; 6% meatal stenosis requiring meatoplasty Intermittent discharge 5/102 ears (4.9%); completely epithelialized 98/102 ears (96.1%); 82/89 ears water resistant (92.1%); 17/102 ears need another op (16.7%)
HA, hydroxyapatite.
of the English-language literature from 1950 to March 31, 2007 was performed using a combination of the key words “cholesteatoma,” “mastoid,” “mastoidectomy,” “middle ear,” “mastoiditis,” “mastoid obliteration,” and “surgical flaps.” The content of each abstract was reviewed in order to identify studies on surgical flaps for mastoid cavity reconstruction. All the chosen articles were read in full, and their references were cross-checked for reports that were not identified from the Medline search. The surgical series identified in Table 3 all involved using nonvitalized bone or alloplastic materials as
fillers. The results from the present study compared extremely favorably with the other reports. The reason for reporting the outcome on primary and secondary obliterations separately is because they represent different surgical categories. This report allows readers to examine our results on discharging mastoid cavities alone. In spite of a dedicated Ear Audit Clinic, 12 ears in the present cohort missed the scheduled yearly follow-up. The recommended statistical methods to calculate the residual or recurrence rate in such a situation are survival analyses. The
912
Otolaryngology–Head and Neck Surgery, Vol 137, No 6, December 2007
advantage of a survival analysis is that it permits the full use of “censored” data up to the time of re-evaluation. There were no recurrent cholesteatomas, but the incidence of residual cholesteatoma at 3 years and 5 years was 97.0% (95% CI: 92.9%, 100.0%) and 91% (95% CI: 83.1%, 100.0%), respectively. For other parameters, cut-off analysis was used for long-term outcome. The 5-year otorrhea rate was zero, and 96% of the cavities became completely epithelialized. Twenty out of 24 ears were water resistant, and 50% had an air-bone gap of 20 dB or better. One can only speculate as to the reasons for our favorable results in comparison to other reports. The senior author had been meticulous in removing diseased mastoid cells from the infected cavity before performing the obliteration. Other obliterating materials had been tried by the senior author before but were eventually abandoned. Bone paste lost volume with time, resulting in the redevelopment of a cavity. Hydroxyapatite cement gave an unacceptable rate of postoperative infection, possibly because it was nonporous and would not allow fluid or blood accumulated beneath it to escape. The cement also did not fare well in an infected environment.17 Porous hydroxyapatite granules offered the best results in the hand of the senior author. The pores inside and between the granules allow ingrowth of bone and fibrous tissue. The authors also credit their success to the mid-temporal flap. Black, the pioneer of this flap, also reported excellent results on his mastoid reconstructions (Table 3). Reports on the use of free grafts or pedicle flaps as tissue coverings for alloplastic fillers or bone paste appeared to have less favorable results (Table 3).
CONCLUSION This study shows that a combination of mid-temporal pericranial flap and inferiorly based pedicled periosteal flaps gives excellent results in mastoid obliteration, even when the cavity is filled with avascular alloplastic materials. Over 95% of the obliterated cavities became completely epithelialized and free of otorrhea, and most ears were water resistant. The outcome compares very favorably to other published reports.
AUTHOR INFORMATION From the Department of Otolaryngology (Dr Yung) and Research and Development, The Ipswich Hospital NHS Trust, Ipswich, UK (Dr Smith) Corresponding author: Matthew Yung, PhD, FRCS, Department of Otolaryngology, The Ipswich Hospital NHS Trust, Heath Road, Ipswich, Suffolk, IP4 5PD, UK. E-mail address: [email protected].
AUTHOR CONTRIBUTIONS Matthew Yung, data collection, literature search, writing manuscript; Phillip Smith, data analysis, assist in writing manuscript.
FINANCIAL DISCLOSURE None.
REFERENCES 1. Symth GDL. Cholesteatoma surgery: the influence of the canal wall. Laryngoscope 1985;95:92– 6. 2. Black B. Mastoidectomy elimination. Laryngoscope 1995;105:1–30. 3. Palva T. Surgery of chronic ear without cavity. Results in 130 cases with musculoperiosteal flap and fasciotympanoplasty. Arch Otolaryngol 1960;77:570 – 80. 4. Kisch J. The use of temporal muscle grafts in the radical mastoid obliteration. J Laryngol Otol 1928;43:856 –9. 5. East CA, Brough MD, Grant HR. Mastoid obliteration with the temporoparietal fascia flap. J Laryngol Otol 1991;105:417–20. 6. Yung MM, Karia KR. Mastoid obliteration with hydroxyapatite—the value of high resolution CT scanning in detecting recurrent cholesteatoma. Clin Otolaryngol 1997;22:553–7. 7. Tos M. Manual of middle ear surgery. Vol 2. Mastoid surgery and reconstructive procedures. Stuttgart, New York: Georg Thieme Verlag; 1995. p. 262-84. 8. Roberson JB, Mason TP, Stidham KR. Mastoid obliteration: autogenous cranial bone paste reconstruction. Otol Neurotol 2003;24:132– 40. 9. D’Arc, MB, Daculsi G, Emam N. Biphasic ceramics and fibrin sealant for bone reconstruction in ear surgery. Ann Otol Rhinol Laryngol 2004;113:711–20. 10. Estrem SA, Highfill G. Hydroxyapatite canal wall reconstruction/ mastoid obliteration. Otolaryngol Head Neck Surg 2000;120:345–9. 11. Ramsey MJ, Merchant SN, McKenna MJ. Postauricular periostealpericranial flap for mastoid obliteration and canal wall down tympanomastoidectomy. Otol Neurotol 2004;25:873– 8. 12. Mercke U. The cholesteatomatous ear one year after surgery with obliteration technique. Am J Otol 1987;8:534 – 6. 13. Symth GDL. Surgical treatment of cholesteatoma. The role of staging in closed operations. Ann Otol Rhinol Laryngol 1988;97:667–9. 14. Vercruysse JP, De Foer B, Pouillon M, et al. The value of diffusionweighted MR imaging in the diagnosis of primary acquired and residual cholesteatoma: a surgical verified study of 100 patients. Eur Radiol 2006;16:1461–7. 15. Jansen VD. Mastoid obliteration with primary ossicular reconstruction. J Otolaryngol 1981;10:321– 4. 16. Leatherman BD, Dornhoffer JL. The use of demineralized bone matrix for mastoid cavity obliteration. Otol Neurotol 2004;25:22– 6. 17. Mahendran S, Yung MW. Mastoid obliteration with hydroxyapatite cement: the Ipswich experience. Otol Neurotol 2004;25:19 –21. 18. Black B. Mastoidectomy elimination: obliterate, reconstruct, or ablate? Am J Otol 1998;19:551–7. 19. Nyrop M, Bonding P. Extensive cholesteatoma: long-term results of three surgical techniques. J Laryngol Otol 1997;111:521– 6. 20. Palva T. Surgical treatment of chronic middle ear disease. Acta Otolaryngol 1988;105:82–9.
ORIGINAL RESEARCH—OTOLOGY
Mid-temporal pericranial and inferiorly based periosteal flaps in mastoid obliteration Matthew Yung, PhD, FRCS, and Phillip Smith, PhD, Ipswich, UK OBJECTIVE: To study the outcome of mastoid obliterations where mid-temporal pericranial flap and inferiorly based periosteal flap were used. STUDY DESIGN: A prospective longitudinal study with a minimum follow-up of 1 year. SUBJECTS AND METHODS: Between 2000 and 2005, 102 mastoid obliterations were performed on 96 patients, the majority having hydroxyapatite granules as fillers. Of these, 31 were primary obliterations and 71 were secondary obliterations. Thirty were on children under 16. RESULTS: Using Kaplan-Meier analysis, the percentage of ears that were free of residual cholesteatoma at 5 years was 91% (95% CI: 83.1%, 100.0%). There was no recurrent cholesteatoma. Cutoff analysis was used to study other parameters at 5 years. The 5-year otorrhea rate was zero, and 96% of the cavities became completely epithelialized. Twenty out of 24 ears were water resistant, and 50% had air-bone gaps of 20 dB or better. CONCLUSION: The outcome of this technique compares favorably to other reports. © 2007 American Academy of Otolaryngology–Head and Neck Surgery Foundation. All rights reserved.
M
astoid obliteration is an established technique for the treatment of chronic infection of the mastoid bone.1 A variety of techniques of mastoid obliteration have been described. These are all based on the same principle: the cavity is reduced using either biological tissues or biocompatible materials as fillers and a soft tissue flap is then used to cover the obliteration.2 For some of these techniques, the filler and the covering material is a single soft tissue flap.3,4 The obliterated cavity must be lined by keratinizing squamous epithelium for it to remain dry and water resistant. The migration of squamous epithelium over the soft tissue flap begins at the remnant of pre-existing meatal skin or tympanic membrane. It could be impeded by a number of unfavorable factors, such as localized infection, ischemia, or necrosis of the flap. Theoretically, a healthy vascularized flap should re-epithelialize faster than an ischemic flap. In general, soft tissue flaps used in mastoid obliteration can be categorized into pedicled flaps (eg, superiorly based musculo-periosteal flap,4 anteriorly based musculo-perios-
teal flap3) or axial flaps (eg, mid-temporal pericranial flap,2 temporoparietal fascial flap5). The senior author has been using a combination of two flaps, an inferiorly based periosteal flap (pedicle flap) and a mid-temporal pericranial flap (axial flap), to line the obliterated cavities since 2000. The present report is a prospective longitudinal study of the ears following mastoid obliteration.
PATIENTS AND METHODS Ear Audit Clinic The senior author has had a policy of a yearly review of patients following mastoid surgery. A weekly ear audit clinic was set up and the senior author personally evaluated the patients at each postoperative follow-up. The clinical information was recorded on a data proforma.
Operative Technique for Mastoid Obliteration A postaural incision was used in all patients to allow exposure and harvesting of the inferiorly based periosteal flap and the mid-temporal flaps. The postaural skin flap was lifted forward to expose the periosteum over the mastoid bone. An inferiorly based periosteal flap was then created with a broad base near the tip of the mastoid process. In most cases, the senior author incorporated the adjacent temporalis fascia into the apex of the periosteal flap to make it longer (Fig 1A). The mid-temporal pericranial flap is based on the midtemporal artery. The flap has been previously described in detail by Black.2 The temporalis muscle overlying the pericranium was dissected off to expose the pericranium. The middle temporal artery has a dual origin from both the deep temporal and the superficial temporal arteries. Besides the main axial vessels, a network of smaller vessels is also present. The middle temporal vessels are usually readily identifiable as a leash of vessels running posterosuperiorly. The flap was fashioned according to the path of the middle temporal vessels and was made as long as possible and at least 40 to 50 mm wide.
Received June 24, 2007; revised August 20, 2007; accepted September 9, 2007.
0194-5998/$32.00 © 2007 American Academy of Otolaryngology–Head and Neck Surgery Foundation. All rights reserved. doi:10.1016/j.otohns.2007.09.014
Yung and Smith
Mid-temporal pericranial and inferiorly based . . .
907
Figure 1 (A) The outline of the inferiorly based periosteal flap. A small piece of temporalis fascia is incorporated into the tip of the flap. zg, groove across the root of the zygoma; smc, supramastoid crest; pf, periosteal flap. (B) The outline of the mid-temporal flap. The inferiorly based periosteal flap is already prepared. zg, groove across the root of the zygoma; mtf, mid-temporal flap; pf, periosteal flap. (C) The mastoid cavity is obliterated with hydroxyapatite granules and cartilage plates. The inferiorly based periosteal flap and the mid-temporal flaps are used to cover the fillers in the mastoid cavity. mtf, mid-temporal flap; pf, periosteal flap.
In most cases, the senior author enlarged the flap by extending it posteriorly and incorporating part of the pericranium overlying the occipital bone into the flap (Fig 1B). This resulted in a big flap that was long enough to reach the annulus and cover the whole attic area. A revision mastoidectomy was then performed to remove disease in the mastoid bowl and eradicate any infected mastoid cells. Healthy-looking epithelial lining of the mastoid cavity was elevated from the underlying bone and used
to reline the ear canal. The three semicircular canals, dura, and the sigmoid sinus were skeletonized to facilitate exenteration of the diseased mastoid cells. Porous hydroxyapatite granules (0.5-1.0 mm diameter; Pentax, Smith & Nephew Richards, Tennesee, USA) were soaked in antibiotic solution before they were put into the mastoid cavity. The bulk of the granules were placed within the mastoid cavity, with only a thin layer put into the epitympanic space.
908
Otolaryngology–Head and Neck Surgery, Vol 137, No 6, December 2007
A large piece of conchal cartilage was harvested and sliced into thin plates using the Kurz Precise cartilage knife (Kurz, Dusslingen, Germany). The cartilage plates were used to tile over the hydroxyapatite granules, starting at the epitympanum. Thicker sheets were used to reconstruct the attic wall, supported behind by the hydroxyapatite granules. Once all the cartilaginous “tiles” were in place, the inferior portion of the reconstructed area was covered by the inferiorly based periosteal flap. The main part of the obliterated area was covered by the mid-temporal pericranial flap, from the epitympanum to the floor of the new canal. Part of the mid-temporal pericranial flap inevitably would lie over the inferiorly based periosteal flap (Fig 1C). Any existing healthy-looking mastoid lining was placed on top of the mid-temporal pericranial flap. For most revision mastoidectomies, a meatoplasty would already be present. Otherwise, a meatoplasty was performed to provide a funnel-shaped opening to the canal.
Cohort of Obliterated Cavities Between 2000 and 2005, a total of 102 mastoid obliterations for chronic otitis media were performed on 96 patients (age range 5-80 years, 64 males and 32 females). A total of 30 obliterations were performed on children under age 16 and six patients had bilateral operations. The minimal follow-up period for each ear was 12 months. Of these 102 mastoid obliterations, 71 were on discharging mastoid cavities (secondary obliteration) and 31 were primary reconstructions of the cavity following a canal wall down mastoidectomy (primary obliteration). Ninety-five cavities had hydroxyapatite granules and cartilage plates as fillers; five smaller cavities only had cartilage as fillers; and two ears had the soft tissue flaps placed directly into the mastoid cavity with no fillers. The patients were followed up at 3 weeks, 6 weeks, 3 months, and 6 months after the operation and then on a yearly basis. At follow-up visits, a microscopic examination of the ear and pure tone audiogram was performed. Patients were encouraged to go swimming to test water resistance. Any otological symptoms from swimming, hair washing, or shower were documented. For ears that were obliterated by hydroxyapatite granules, an interval CT scan of the mastoid bone was routinely performed at 12 months.
Audit Information collected for the present study included the time interval for the ear to become epithelialized, incidence of postoperative otorrhea, patient perception of water-resistant properties, residual or recurrent cholesteatoma, and hearing results. Residual cholesteatoma usually took the form of an epithelial pearl due to remnants of the cholesteatoma matrix left behind the tympanic membrane or within the mastoid bone. The interval CT scan results at 12 months and any subsequent ear operations were also recorded. The residual cholesteatoma rates were studied using KaplanMeier analysis (MedCalc for Windows). Cut-off analysis
was used to study the other surgical outcomes at 5 years, including hearing outcomes. The outcome of primary obliteration and secondary obliteration is compared. At the outset of this study, an audit approach was deemed to be most appropriate. Consequently Institutional Review Board or UK equivalent of approval was not sought. The conduct of this audit was consistent with UK best practice.
RESULTS With the benefit of a dedicated ear audit clinic, 84 of the 96 patients (90 of the 102 ears) attended all the scheduled follow-up visits. A total of 101 ears had been followed up for at least 1 year, 73 ears for 2 years, 48 ears for 3 years, 33 ears for 4 years, and 24 ears for 5 years or longer. Of the 102 ears, 90 had an interval CT scan at 12 months. The purpose of the CT scan was to identify areas of low CT density (suggestive of residual cholesteatoma pearls) in a background of high CT density (hydroxyapatite granules).6 Six ears did not have CT scanning because hydroxyapatite granules were not used in their obliterations. It was felt that the differences in CT densities between residual cholesteatoma pearls and surrounding soft tissues did not give sufficient contrast for little pearls to show up on the scan. For the other six ears, five missed the scan and one patient had left the country. There was no recurrent cholesteatoma in the present cohort, but an attic cavity redeveloped in one ear due to loss of the hydroxyapatite granules in the epitympanum. The attic region was successfully reobliterated. Residual cholesteatomas were detected in five ears; three were in the middle ear (detected at 24, 40, and 51 months postobliteration) while the other two residual cholesteatomas were in the epitympanum (detected 32 and 66 months postobliteration). All the residual cholesteatomas were detected during routine follow-up and appeared as pearls behind the tympanic membrane or the cartilaginous attic wall. None of the residual cholesteatomas were picked up by the interval CT scans at 12 months. Of the 90 interval CT scans (12 months postobliteration), only one showed a shadow beneath the hydroxyapatite granules. This ear was re-explored, and the shadow was found to be due to fibrous tissue rather than cholesteatoma. Using Kaplan-Meier analysis, the percentages of ears within the present cohort that were free of residual cholesteatoma at 3 years and 5 years were 97.0% (95% CI: 92.9%, 100.0%) and 91% (95% CI: 83.1%, 100.0%), respectively. Table 1 depicts the outcome parameters of the ears in the present cohort. The outcome of primary obliterations is compared to that of secondary obliterations. Overall, 98 of the 102 ears (96%) had complete re-epithelialization of the cavity, and over 80% of the ears achieved that within 12 weeks following obliteration. A total of five ears (4.9%) suffered from intermittent otorrhea following obliterations including two incidences
Yung and Smith
Mid-temporal pericranial and inferiorly based . . .
909
Table 1 Outcome of using combined mid-temporal and inferiorly based periosteal flaps as soft tissue covering in mastoid obliteration Primary obliteration (n⫽31)
Parameters
Secondary obliteration (n⫽71)
Total cohort (n⫽102)
5-y cut-off analysis (n⫽24)
Children : adults Time to become fully epithelialized (weeks) Obliterate using HA granules Follow-up periods (months)
17 : 14 Median: 8 wks; 30/31 within 12 wks (96.8%) 27 ears
13 : 58 31 : 71 7 : 17 Median: 8 wks; 56/71 Median: 8 wks; 86/102 Median: 8 wks; 23/ 24 within 12 wks within 12 wks within 12 wks (95.8%) (84.3%) (78.9%) 68 ears 95 ears 22 ears
Median: 32 mos; range: 12-75 mos
Median: 34 mos; range: 12-80 mos
Median: 67 mos; range: 60-80 mos
12-month interval CT scan Residual cholesteatoma Recurrent cholesteatoma Postoperative otorrhea
27/31 completed
Median: 35 mos; range: 12-80 mos (1 immigrated at 8 mos) 63/71 completed
90/102 completed
21/24 completed
0 children, 2 adults
2 children, 1 adult
2 children, 3 adults
2 children, 2 adults
0
0
0
0
0
Single episode: 1; intermittent: 5 (4.9%)
0
Appearance of cavity
30 complete epith. (96.8%); 1 incomplete epith. 27 swim/hair wash (87.1%); 2 unknown
Single episode: 1; intermittent: 5 (6.9%) (2 myringitis; 1 retraction pocket; 1 meatal stenosis; 1 incomplete epith.) 68 complete epith. (95.8%); 3 incomplete epith. 51 swim/hair wash ⫹ 4 hair wash (77.5%); 11 unknown 2 incomplete epith. ⫹ 1 meatal narrowing ⫹ 2 myringitis (7.0%)
Water resistance
Not water resistant
1 incomplete epith. ⫹ 1 TM perforation (6.5%)
Subsequent ear operation
3 second-look ⫹ 1 myringoplasty ⫹ 2 removal of residual cholest. (19.4%)
2 second-look ⫹ 1 myringoplasty ⫹ 3 removal of residual cholest. ⫹ 1 repair retraction pocket ⫹ 2 meatoplasties ⫹ 1 ossiculoplasty ⫹ 1 skin grafting of cavity (15.5%)
98 complete epith. (96.1%); 4 incomplete epith. 82 water-resistant (80.0%); 13 unknown
23 complete epith. (95.8%); 1 incomplete epith. 19 swim/hair wash ⫹ 1 hair wash (83.3%); 1 unknown 7 non-water-resistant 1 incomplete epith. (6.9%) ⫹ 1 TM perforation ⫹ 1 myringitis (12.5%) 3 second-look ⫹ 2 5 second-look ⫹ 2 myringoplasties myringoplasties ⫹ 5 ⫹ 4 removal of removal of residual residual cholest. cholest. ⫹ 1 repair ⫹1 retraction pocket ⫹ ossiculoplasty ⫹ 2 meatoplasties ⫹ 1 1 skin graft of ossiculoplasty ⫹ 1 cavity (45.8%) skin grafting of cavity (16.7%)
HA, hydroxyapatite; TM, tympanic membrane; epith., epithelialization; cholest, cholesteatoma; mos, months; wks, weeks.
of myringitis and one incidence for each of the following: attic retraction pocket, meatal stenosis, and incomplete epithelialization of the obliterated cavity. Most patients tried swimming or washing their hair without ear protection to test water resistance; 81 ears were
water resistant while 7 were not. The reasons for the non– water resistance are given in Table 1. The water-resistant property was unknown in 13 ears because of coexisting pathology in the opposite ear, or patients’ reluctance to change the old habit of wearing ear plugs.
910
Otolaryngology–Head and Neck Surgery, Vol 137, No 6, December 2007
A total of 17 ears had further surgery during the observation period for a variety of reasons. They were secondlook tympanoplasties (five ears), surgery to remove residual cholesteatoma pearls (five ears), myringoplasties (two ears), meatoplasties (two ears), ossiculoplasty (one ear), split skin grafting of nonepithelialized areas in the cavity (one ear), and repair of attic retraction pocket (one ear). None of the patients had major postoperative complications, such as dead ear, facial palsy, or prolonged dizziness. Trimming of the soft tissue flap was performed on three ears because of tip necrosis. It is recognized that outcome reports on cholesteatoma surgery should be based on long-term studies. Therefore a cut-off analysis on the subset of ears that had a minimum follow-up period of 5 years is included in Table 1. The subset comprised 24 ears, with a median follow-up period of 67 months. The dry ear rate was 100%, with 96% of the obliterated cavities completely re-epithelialized. Also, 83% of the ears in this subset were water resistant. The presentation of the hearing results is less straightforward. The authors decided to present the 5-year hearing results to show what has been achieved, accepting that 11 out of 24 ears had more than one operation on the ear (Table 2). The average 5-year air-bone gap was 22.29 dB (95% CI: ⫺2.33 dB, 46.92 dB). Two patients had worsening of the air-bone gaps. Overall, 5-year air-bone gaps of 20 dB or better were achieved in 50% of the ears.
DISCUSSION A number of soft tissue flaps have been described for mastoid obliteration.2 The choice is often down to personal preference, rather than based on clinical evidence. When a soft tissue flap is placed over living bone, it gets nourishment directly from the bone. Indeed, free temporalis fascia graft is commonly used by surgeons to line the mastoid cavity.7 When a soft tissue flap is placed over nonvascular materials, such as bone paste, hydroxyapatite, or titanium canal prosthesis, it needs a pedicle or feeding vessels for nourishment. When free grafts are placed over nonvascular materials, they could break down and lead to exposure of the underlying materials or otorrhea.8-10
Table 2 Five-year hearing results following mastoid cavity obliteration (nⴝ24) Number of ears (%) Air-bone gap
Preoperative
0-10 dB 11-20 dB 21-30 dB ⬎30 dB
1/24 4/24 8/24 11/24
(4.2%) (16.7%) (33.3%) (45.8%)
Postoperative 5/24 7/24 5/24 7/24
(20.8%) (29.2%) (20.8%) (29.2%)
The senior author has been using hydroxyapatite granules as fillers in mastoid obliteration since 1990. Initially, the granules were covered by just an inferiorly based periosteal flap. It was found that the flap was not long enough to reach the annulus and the attic area. The senior author modified his techniques since 2000. Cartilage slices were used to cover the hydroxyapatite granules to smooth out the crevices and reinforce the attic reconstruction. In addition to the inferior-based periosteal flap, a superiorly based midtemporal pericranial flap was also used to provide vascular covering of the cavity. The present report is an observational study on 102 ears using this technique. The senior author favors the mid-temporal pericranial flap because of its good blood supply. It is, however, thin and can only be used as a covering flap rather than as filler. The reason for using two flaps is to provide extra cover over the granules. It is not yet clear how dependent the pericranial flap is on the middle temporal artery. Ramsey et al11 used an inferiorly based periosteal-pericranial flap to cover bone paste in mastoid obliteration and still reported a 65% complete dry ear rate. That indicates that the nourishment of the pericranium is not totally dependent on the middle temporal vessels. A concern with mastoid obliteration is that residual cholesteatomas may be buried underneath the fillers. Secondlook operation12 has been suggested as a possible way to detect these residuals. However, it has been reported that many ears developed residual cholesteatoma even when they have been cleared at the second-look operation.13 In the present study, the residual cholesteatoma only became apparent several years after the obliteration. None were detected by interval scanning at 12 months. It may well be that scanning of the obliterated ears at 12 months is too early. The authors are currently considering having the interval scan scheduled at 24 months instead. There is emerging evidence that MRI scan can differentiate cholesteatoma from other soft tissues within the mastoid bone but is still not specific enough to replace a second-look operation.14 Whatever policy one uses to detect residual cholesteatoma, there is an almost universal agreement that long-term follow-up of these patients is necessary.15,16 The aim of mastoid obliteration is to stop otorrhea and render the ear water resistant. In the present cohort, 95% of the ears were free of otorrhea, and 96% of the cavities were completely re-epithelialized. For the ears with otorrhea resulting from myringitis or incomplete epithelialization, they were treated with 2% acetic acid eardrops or sprays, and topical antiobiotic-steroid eardrops only if necessary. For the patients who had persistent otorrhea, revision surgeries were offered, such as skin grafting of the nonepithelialized areas. In the present cohort, over 80% of the ears were completely re-epithelialized within 12 weeks following the obliteration. Eighty-two out of 89 ears were found to be water resistant, and most patients could enjoy water sports. In order to facilitate the comparison of our results to other reports, a literature review was performed. A Medline search
Yung and Smith
Mid-temporal pericranial and inferiorly based . . .
911
Table 3 Literature review
Author
Size of cohort
Follow-up periods
Fillers
Covering tissue
Black18
124
Average 3 years
HA canal prosthesis
Mid-temporal flap
Roberson8
62
0.2-54.8 months
Bone paste
D’Arc9
72
1-158 months; average 46 months
Biphasic calcium phosphate granules
Free temporalis fascia Free temporalis fascia
Estrem10
31
1-74 months
HA canal prosthesis; HA granules
Free homograft dura ⫹ temporalis fascia
Leatherman16
11
6-20 months
Demineralized bone matrix ⫹ cartilage
Nyrop19
27
10-13 years
Palva20
185
1-11 years; average 5 years
Bone chip and paste Bone chip and paste
Free perichondrium graft ⫹ Palva flap Muscle flap
Ramsey11
60
12-80 months; average 32 months
Bone paste
Inferiorly based periostealpericranial flap ⫹ split skin
Present series
102
12-80 months; median 34 months
HA granules ⫹ cartilage
Mid-temporal flap ⫹ inferiorly based periosteal flap
Palva flap
Outcome Implant exposure 2.5%; avascular necrosis 3%; infection 4%; myringitis 8% 8/62 still discharging (13.8%) 19% still otorrhea at 12 months (a number of patients had leakage of granules) HA canal wall 25% failure; HA granules 13% failure (failure include extrusion of HA, and granulation tissue on the lining) 2/11 prolonged granulation, thought to be due to exposed bone matrix 89% dry ear; average no. of op ⫽ 1.2 10% had revision surgery; 80% remained dry 10% persisent discharge; 25% intermittent discharge; 6% meatal stenosis requiring meatoplasty Intermittent discharge 5/102 ears (4.9%); completely epithelialized 98/102 ears (96.1%); 82/89 ears water resistant (92.1%); 17/102 ears need another op (16.7%)
HA, hydroxyapatite.
of the English-language literature from 1950 to March 31, 2007 was performed using a combination of the key words “cholesteatoma,” “mastoid,” “mastoidectomy,” “middle ear,” “mastoiditis,” “mastoid obliteration,” and “surgical flaps.” The content of each abstract was reviewed in order to identify studies on surgical flaps for mastoid cavity reconstruction. All the chosen articles were read in full, and their references were cross-checked for reports that were not identified from the Medline search. The surgical series identified in Table 3 all involved using nonvitalized bone or alloplastic materials as
fillers. The results from the present study compared extremely favorably with the other reports. The reason for reporting the outcome on primary and secondary obliterations separately is because they represent different surgical categories. This report allows readers to examine our results on discharging mastoid cavities alone. In spite of a dedicated Ear Audit Clinic, 12 ears in the present cohort missed the scheduled yearly follow-up. The recommended statistical methods to calculate the residual or recurrence rate in such a situation are survival analyses. The
912
Otolaryngology–Head and Neck Surgery, Vol 137, No 6, December 2007
advantage of a survival analysis is that it permits the full use of “censored” data up to the time of re-evaluation. There were no recurrent cholesteatomas, but the incidence of residual cholesteatoma at 3 years and 5 years was 97.0% (95% CI: 92.9%, 100.0%) and 91% (95% CI: 83.1%, 100.0%), respectively. For other parameters, cut-off analysis was used for long-term outcome. The 5-year otorrhea rate was zero, and 96% of the cavities became completely epithelialized. Twenty out of 24 ears were water resistant, and 50% had an air-bone gap of 20 dB or better. One can only speculate as to the reasons for our favorable results in comparison to other reports. The senior author had been meticulous in removing diseased mastoid cells from the infected cavity before performing the obliteration. Other obliterating materials had been tried by the senior author before but were eventually abandoned. Bone paste lost volume with time, resulting in the redevelopment of a cavity. Hydroxyapatite cement gave an unacceptable rate of postoperative infection, possibly because it was nonporous and would not allow fluid or blood accumulated beneath it to escape. The cement also did not fare well in an infected environment.17 Porous hydroxyapatite granules offered the best results in the hand of the senior author. The pores inside and between the granules allow ingrowth of bone and fibrous tissue. The authors also credit their success to the mid-temporal flap. Black, the pioneer of this flap, also reported excellent results on his mastoid reconstructions (Table 3). Reports on the use of free grafts or pedicle flaps as tissue coverings for alloplastic fillers or bone paste appeared to have less favorable results (Table 3).
CONCLUSION This study shows that a combination of mid-temporal pericranial flap and inferiorly based pedicled periosteal flaps gives excellent results in mastoid obliteration, even when the cavity is filled with avascular alloplastic materials. Over 95% of the obliterated cavities became completely epithelialized and free of otorrhea, and most ears were water resistant. The outcome compares very favorably to other published reports.
AUTHOR INFORMATION From the Department of Otolaryngology (Dr Yung) and Research and Development, The Ipswich Hospital NHS Trust, Ipswich, UK (Dr Smith) Corresponding author: Matthew Yung, PhD, FRCS, Department of Otolaryngology, The Ipswich Hospital NHS Trust, Heath Road, Ipswich, Suffolk, IP4 5PD, UK. E-mail address: [email protected].
AUTHOR CONTRIBUTIONS Matthew Yung, data collection, literature search, writing manuscript; Phillip Smith, data analysis, assist in writing manuscript.
FINANCIAL DISCLOSURE None.
REFERENCES 1. Symth GDL. Cholesteatoma surgery: the influence of the canal wall. Laryngoscope 1985;95:92– 6. 2. Black B. Mastoidectomy elimination. Laryngoscope 1995;105:1–30. 3. Palva T. Surgery of chronic ear without cavity. Results in 130 cases with musculoperiosteal flap and fasciotympanoplasty. Arch Otolaryngol 1960;77:570 – 80. 4. Kisch J. The use of temporal muscle grafts in the radical mastoid obliteration. J Laryngol Otol 1928;43:856 –9. 5. East CA, Brough MD, Grant HR. Mastoid obliteration with the temporoparietal fascia flap. J Laryngol Otol 1991;105:417–20. 6. Yung MM, Karia KR. Mastoid obliteration with hydroxyapatite—the value of high resolution CT scanning in detecting recurrent cholesteatoma. Clin Otolaryngol 1997;22:553–7. 7. Tos M. Manual of middle ear surgery. Vol 2. Mastoid surgery and reconstructive procedures. Stuttgart, New York: Georg Thieme Verlag; 1995. p. 262-84. 8. Roberson JB, Mason TP, Stidham KR. Mastoid obliteration: autogenous cranial bone paste reconstruction. Otol Neurotol 2003;24:132– 40. 9. D’Arc, MB, Daculsi G, Emam N. Biphasic ceramics and fibrin sealant for bone reconstruction in ear surgery. Ann Otol Rhinol Laryngol 2004;113:711–20. 10. Estrem SA, Highfill G. Hydroxyapatite canal wall reconstruction/ mastoid obliteration. Otolaryngol Head Neck Surg 2000;120:345–9. 11. Ramsey MJ, Merchant SN, McKenna MJ. Postauricular periostealpericranial flap for mastoid obliteration and canal wall down tympanomastoidectomy. Otol Neurotol 2004;25:873– 8. 12. Mercke U. The cholesteatomatous ear one year after surgery with obliteration technique. Am J Otol 1987;8:534 – 6. 13. Symth GDL. Surgical treatment of cholesteatoma. The role of staging in closed operations. Ann Otol Rhinol Laryngol 1988;97:667–9. 14. Vercruysse JP, De Foer B, Pouillon M, et al. The value of diffusionweighted MR imaging in the diagnosis of primary acquired and residual cholesteatoma: a surgical verified study of 100 patients. Eur Radiol 2006;16:1461–7. 15. Jansen VD. Mastoid obliteration with primary ossicular reconstruction. J Otolaryngol 1981;10:321– 4. 16. Leatherman BD, Dornhoffer JL. The use of demineralized bone matrix for mastoid cavity obliteration. Otol Neurotol 2004;25:22– 6. 17. Mahendran S, Yung MW. Mastoid obliteration with hydroxyapatite cement: the Ipswich experience. Otol Neurotol 2004;25:19 –21. 18. Black B. Mastoidectomy elimination: obliterate, reconstruct, or ablate? Am J Otol 1998;19:551–7. 19. Nyrop M, Bonding P. Extensive cholesteatoma: long-term results of three surgical techniques. J Laryngol Otol 1997;111:521– 6. 20. Palva T. Surgical treatment of chronic middle ear disease. Acta Otolaryngol 1988;105:82–9.