Available online at www.sciencedirect.com

American Journal of Otolaryngology–Head and Neck Medicine and Surgery 29 (2008) 363–366www.elsevier.com/locate/amjoto

Original contributions

Minimally invasive inlay and underlay tympanoplastyWen-Hung Wang, MD, Yen-Chun Lin, MD⁎

Department of Otolaryngology—Head and Neck Surgery, Chang Gung Memorial Hospital, Pu-Tzu City, Chiayi County, TaiwanGraduate Institute of Clinical Medical Sciences, Chang Gung University College of Medicine, No. 123, Dapi Road, Niaosong Township,

Kaohsiung County, Taiwan

Received 27 August 2007

Abstract Purpose: The objective of this study is to understand the outcome of minimally invasive topical

⁎ CorrespondingMemorial Hospital anWest Sec, Chia-PuTel.: +886 5 3621000

E-mail address: sa

0196-0709/$ – see frodoi:10.1016/j.amjoto.2

anesthetized transcanal inlay and underlay tympanoplasty, and to compare these 2 procedures inhearing result, take rate, perioperative pain, and operation time.Materials and methods: This is a retrospective study conducted from September 2003 to December2006. Forty-eight tympanoplasty in 46 patients, 28 inlay and 20 underlay procedures, with small- tomedium-sized tympanic membrane perforations without middle ear lesion, were studied in a tertiaryreferral center. Interventions included otologic examination, perioperative hearing evaluation, localanesthetized transcanal inlay, or underlay tympanoplasty. The outcome measurements were thefollowing: the take rate and audiometric result at the last follow-up visit, perioperative pain, and durationof surgery. The statistical methods used were t test, Mann-Whitney U test, χ2 test, and Fisher exact test.Results: The take rate were 82.1% in the inlay group and 85% in the underlay group, withoutsignificant difference (P = .79) at the last follow-up visit. Air-bone gap closure was 6.3 ± 2.5 dB inthe inlay group and 9.3 ± 3.2 dB in the underlay (P = .07). Linear analogue scale of perioperativepain was lower in the inlay group with significance (1.7 ± 1.2 in the inlay and 4.6 ± 1.9 in theunderlay group, P b .001). The duration of the surgery was significantly shorter in the inlay group(31.8 ± 13.9 minutes for the inlay group and 75.9 ± 14.6 minutes for the underlay group, P b .001).Conclusions: Minimally invasive topically anesthetized transcanal tympanoplasty providessatisfactory surgical and audiometric outcome both in inlay and underlay procedures. The takerate and extent of hearing recovery are similar in both groups. However, inlay tympanoplasty issuperior to transcanal tympanoplasty because of less discomfort and shorter operation time.

© 2008 Elsevier Inc. All rights reserved.

1. Introduction

The 2 procedures used to repair perforation of thetympanic membrane are the overlay and the underlaytechnique. The most accepted and frequently used techniquegenerally is underlay of temporalis fascia. The underlayprocedure could be done through either postauricular ortranscanal approach. However, general anesthesia is usuallyrequired in the temporalis fascia underlay tympanoplasty.

author. Department of Otolaryngology, Chang Gungd Chang Gung University College of Medicine, 6,Road, Pu-Tzu City, Chiayi County 613, Taiwan.; fax: +886 5 3623048.rah_travel@hotmail.com (Y.-C. Lin).

nt matter © 2008 Elsevier Inc. All rights reserved.007.11.002

In 1998, Roland Eavey [1] developed a new inlaytechnique for tympanoplasty using a cartilage graft througha transcanal approach, which had several practical advan-tages. The procedure can be done in an office setting underlocal anesthesia; no postauricular incision is required; thereis no need for external canal packing or middle ear supportto stabilize the graft; the procedure is less expensivebecause of diminished operative and recovery room time[1]. The take rate and audiometric results after inlaycartilage tympanoplasty is similar to underlay tympano-plasty. The take rate is claimed more than 83.3% to 100%[1-8], and the postoperative hearing results are generallygood [3].

In our institute, before May 2005, the protocol ofminimally invasive tympanoplasty for perforated tympanic

364 W.-H. Wang, Y.-C. Lin / American Journal of Otolaryngology–Head and Neck Medicine and Surgery 29 (2008) 363–366

membrane repair was transcanal underlay tympanoplasty.The rational of choosing transcanal underlay tympanoplastyas the standard protocol is because the grafting material,temporalis fascia, is physiologically compatible with thetympanic membrane. Moreover, the underlay procedure iswidely accepted as feasible and reliable. The procedure wasconducted under local anesthesia in all cases.

We collected a cohort of consecutive 20 patients receiving20 procedures. We found the transcanal underlay tympano-plasty achieved satisfactory surgical and audiometric out-comes. However, the duration of surgery was long and somepatients complained of discomfort during the operation. Wetherefore changed our protocol and used inlay procedureafter May 2005.

The aim of the current study was to compare the results, interms of hearing improvement, take rate, perioperative pain,and operation time, of minimally invasive transcanal inlayand transcanal underlay tympanoplasty.

2. Methods

2.1. Patients

A total of 46 patients aged 16 to 87 years who met thecriteria were operated on by a single surgeon duringSeptember 2003 to December 2006. Only patients whomet the following criteria received the minimally invasiveprocedure. The inclusion criteria were (1) small- to medium-sized perforations, (2) nonmarginal perforation, (3) sufficientexternal canal width with clear view of all perforatedmargins, (4) dry ear without middle ear disease or concurrentexternal ear infection, and (5) at least 3 months fornonhealing of the perforations. The demographic character-istics of patients were listed in Table 1. Twenty patientsenrolled from September 2003 to May 2005 underwentunderlay tympanoplasty, and 26 patients enrolled from May2005 to December 2006 underwent inlay tympanoplasty.There were no differences in age and sex distribution in these2 groups.

Table 1Baseline characteristics of the patients according to intervention orcontrol group

Inlay technique,n = 28 ears/26 patients

Underlay technique,n = 20 ears/20 patients

P

SexMale 10 (38.5%) 7 (35%) .37Female 16 (61.5%) 13 (65%)

Age (y)Mean ± SD 56.0 ± 15.3 54.2 ± 15.5 .72

Bilateralperforation

2 0

Previoustympanoplasty

4 (15.4%) 2 (10%) .83

Meanfollow-up (mo)

11.2 ± 2.7 15.9 ± 3.4

2.2. Technique

The inlay tympanoplasty procedure carried out in thisstudy was a modified technique, described by Lubianca-Neto[2]. The tragal cartilage has been harvested with perichon-drium preserved on both surfaces. A Beaver #6700miniblade scores the entire cartilage edge circumferentiallybetween the layers of perichondrium. The edges curl apart,creating an appearance similar to the wings of a butterfly.The butterfly cartilage graft is grasped with a toothlessalligator forceps and inserted in a transcanal fashion. Nosplit-thickness skin graft was placed over the perichondriumof the cartilage graft at the conclusion of the case.

The underlay tympanoplasty procedure carried out wasdescribed by Rizer [9]. Both of the procedures wereconducted under local anesthesia, and patients were preparedand draped for a sterile procedure. Topical anesthesia wasinjected subcutaneously with lidocaine and epinephrine1:100000 solutions into the external auditory canal, themeatal surface of the tragus for inlay tympanoplasty and theposterior-superior portion of the postauricular sulcus forunderlay tympanoplasty. The tympanic membrane perfora-tion was viewed and approached transcanally under themicroscope. No external ear canal packing was done in theinlay procedure. Mastoid dressing was applied in underlaycases for better compression of the postauricular wound.

2.3. Outcome measurements

The primary end points were the take rate and the hearinglevels at the last follow-up visit. For hearing results, thepostoperative air-bone gap (ABG) was grouped as no gap,equal to or less than 10 dB, equal to 11 dB and less than20 dB, and equal to or greater than 21 dB. The ABG isdefined as the mean of gaps at frequencies of 500, 1000, and2000 Hz. The change in ABG was measured as preoperativeABG minus postoperative ABG. The secondary end pointswere the actual time of surgery and perioperative pain.Patients were asked to score their pain on a linear analoguescale right after the operation. Perioperative pain was rankedaccordingly: 0 = without pain; 1 to 3 = mild pain; 4 to 5 =moderate pain; 7 to 9 = severe pain; and 10 = the worstpossible pain. The actual time of surgery was measured fromthe first skin incision to the end of external ear dressing.

2.4. Statistical analysis

Data were statistically analyzed using t test, Mann-Whitney U test, Pearson χ2 test, and Fisher exact test (SPSSsoftware, version 12.0; SPSS, Chicago, IL). Significancewas set at P b .05.

3. Results

The take rate in the inlay tympanoplasty group and theunderlay tympanoplasty group were 82.1% and 85%,respectively, on the last follow-up visit (P = .79). The

Table 2Primary end points according to intervention or control group

Inlay technique,n = 28 ears/26 patients

Underlay technique,n = 20 ears/20 patients

P

Take rate atlast follow-up

23/28 (82.1%) 17/20 (85%) .79

Change in the ABG (dB)b0 2 (7.1%) 2 (10%)0–10 18 (64.3%) 9 (45%)11–20 7 (25%) 7 (35%)N21 1 (3.6%) 2 (10%)Mean ofABG closure(mean ± SD)

6.3 ± 2.5 dB 9.3 ± 3.2 dB .07

Table 4Summary of inlay butterfly cartilage tympanoplasty

Author Publishedyear

Ears no.(patients no.)

Type ofanesthesia

Approach Takerate (%)

Eavey [1] 1998 Inlay 9 (11) GA Transcanal 100Lubianca-Neto [2]

2000 Inlay 20 (20) LA Transcanal 90

Mauriet al [3]

2001 Inlay 40 (34) LA Transcanal 88.2 a

(85.3b)Underlay 46(36)

GA NA 86.1a

(83.3 b)Anandet al [6]

2002 Inlay 20 (20) GA or LA Transcanal 90

Couloigneret al [4]

2005 Inlay 59 (51) LA [49]or GA [2]

Transcanal(58 ears) orendaural(1 ears)

71

Underlay 29(26)

GA NA 83

Effat [5] 2005 Inlay 28 (21) GA Transcanal 43Underlay 23(23)

GA Endaural 83

Ghanemet al [7]

2006 Inlay 99 (90) GA Postauricular 92

Sakagamiet al [8]

2006 Underlay391

LA Transcanal 77.7 c

Our study Inlay 28 (26) LA Transcanal 82.1Underlay 20(20)

LA Transcanal 85

Abbreviations: GA, general anesthesia; LA, local anesthesia; NA, notavailable.a Take rate at 30th postoperative day.b Take rate at the last of follow-up visit.c Initial attempt, 304 (77.7%) of 391; reclosure for unsuccessful cases, 70

(80.5%) of 87; overall rate, 374 (95.7%) of 391.

365W.-H. Wang, Y.-C. Lin / American Journal of Otolaryngology–Head and Neck Medicine and Surgery 29 (2008) 363–366

mean follow-up periods were 11.2 ± 2.7 and 15.9 ±3.4 months in the inlay and underlay group. The take ratedid not differ between groups at the last follow-up visit(Table 2). After inlay procedure, the recurrent perforationrate was 17.9% (n = 5) because of infection for 3 and graftdisplaced for 2. After underlay procedure, failure rate was15% (n = 3) because of infection for all. However, recurrentperforations were all smaller than the initial perforations, and“wait and see” policy was adopted.

The mean ABG closures in the inlay and underlay groupswere 6.3 ± 2.5 and 9.3 ± 3.2 dB without significantdifference (P = .07). In the inlay group, there was closure ofthe ABG to within 10 dB in 71.4% and to within 20 dB in96.4%. In the underlay group, there was closure of the ABGto within 10 dB in 55% and to within 20 dB in 90% (Table 2).

Pain was reported by 21 (80.8%) patients in the inlaygroup and by 20 (100%) patients in the underlay group rightafter the operation. Table 3 revealed that most patients in theinlay group ranked pain as mild, whereas most of the patientsin the underlay group ranked pain as moderate. The meanscore of pain were 1.7 ± 1.2 and 4.6 ± 1.9 in the inlay andunderlay group (P b .001). In the inlay tympanoplasty group,

Table 3Secondary end points according to intervention or control group

Inlay technique,n = 28ears/26 patients

Underlay technique,n = 20ears/20 patients

P

Ranked pain on linear analogue scale0 = without pain 5 01–3 = mild pain 21 44–6 = moderate pain 2 137–9 = severe pain 0 310 = the worstpossible pain

0 0

Mean pain score(mean ± SD)

1.7 ± 1.2 4.6 ± 1.9 b.001⁎

Actual time ofsurgery (min)(mean ± SD)

31.8 ± 13.9 75.9 ± 14.6 b.001⁎

⁎ Significantly different when P b .05.

patients reported mild pain (bursting sound) when insertingthe cartilage. Most of the patients in the underlay groupreported pain when the tympanomeatal flap was developedand approached the bony annulus of the tympanic membrane.

The duration of the surgery was 31.8 ± 13.9 minutes forthe inlay tympanoplasty group and 75.9 ± 14.6 minutes forthe underlay tympanoplasty group (P b .001) (Table 3).

4. Discussion

This is the first study comparing 2 types of minimallyinvasive tympanoplasty procedures under local anesthesia.Patients must fulfill a certain criteria, for example, properwidths of the ear canal and simple central eardrumperforation, to receive minimally invasive procedures. Theadvantages of this retrospective study are that (1) thepatients/disease characteristics are homogenous, and (2)the procedures were assigned randomly.

This study shows good clinical outcomes, in terms of takerate and hearing recovery, in both inlay and underlay groups.From previous study, the take rates are varied from 43% to100% and 83% to 95.7% in inlay and underlay tympano-plasty (Table 4). It is worthy of notice that Effat [5] reported a

366 W.-H. Wang, Y.-C. Lin / American Journal of Otolaryngology–Head and Neck Medicine and Surgery 29 (2008) 363–366

very low take rate (43%) in his inlay tympanoplasty group.He believed it is because of different pathophysiologiceffects of cartilage graft. However, the medium take rate ofinlay tympanoplasty reported is 90% (Table 2, from thereport of Anand et al [6]). Our results are compatible toprevious reported data (82.1% and 85% in inlay andunderlay groups). We indeed have gone through the learningcurve in our first few inlay tympanoplasty cases. Difficultiesin adjusting the size of grafting and cutting cartilage edges ina continuous butterfly fashion are the main problems.

In spite of the stiffness of the cartilage graft, the hearingresults were satisfactory. The hearing recovery is evaluatedby closure of ABG in this study. Both inlay and tympano-plasty close the ABG by 6.3 ± 2.5 and 9.3 ± 3.2 dB.Conceptually, the thinner grafting material provides bettervibration and transmission of sound. However, Dornhoffer[10] showed that there is no difference in postoperativehearing results when comparing cartilage tympanoplasty andperichondrium tympanoplasty. There is also no difference inpostoperative hearing results regarding the size of cartilageused. Mauri et al [3] and Gerber et al [11] have also shownthat cartilage does not affect sound transmission.

Inlay technique is superior to the underlay procedureswhen comparing through the secondary outcome, patientcomfort, and satisfaction. The inlay tympanoplasty is morecomfortable for the patient. In our study, both the inlay andunderlay procedures were performed under local anesthesia.Different levels of pain were reported during the operations.Some patients in the inlay tympanoplasty group complainedof a “pop” sound and mild pain (or discomfort) when thecartilage graft was introduced. Some patients in the underlaygroup complained of pain when the tympanomeatal flap waselevated and tympanic membrane annulus was reached; afew of them need further topical anesthesia.

In addition, we conducted both inlay and underlaytympanoplasty under local anesthesia. Patients were awakeand aware of the entire operation procedures. The length ofactual operation time indeed affects the comfortableness ofpatients. The actual time of operation is doubled in theunderlay group (75.9 ± 14.6 vs 31.8 ± 13.9 minutes);therefore, patients experienced less discomfort in the inlaygroup when compared with the underlay group. The resultsare similar to that of the study by Mauri et al [3]. Theycompared the pain in the postoperative period in the inlayand underlay groups. Despite introducing general anesthesia

procedure in the underlay tympanoplasty group, the patientsof underlay group ranked higher pain when comparing withthe inlay group.

There are several advantages of the inlay group: (1) theincision is limited to tragus, only one incision is needed, (2)no ear canal incision is required, (3) no need to develop thetympanomeatal flap, (4) neither external canal packing normastoid dressing is required, and (5) immediate self-report ofthe hearing change is possible.

5. Conclusion

The take rate and audiometric results after inlaytympanoplasty or underlay tympanoplasty were similar andsatisfactory. However, patients experienced less pain andshorter operation time in the inlay group during operation.The inlay tympanoplasty warrants better option for mini-mally invasive tympanoplasty procedures.

References

[1] Eavey RD. Inlay tympanoplasty: cartilage butterfly technique.Laryngoscope 1998;108(5):657-61.

[2] Lubianca-Neto JF. Inlay butterfly cartilage tympanoplasty (Eaveytechnique) modified for adults. Otolaryngol Head Neck Surg 2000;123(4):492-4.

[3] Mauri M, Lubianca-Neto JF, Fuchs SC. Evaluation of inlay butterflycartilage tympanoplasty: a randomized clinical trial. Laryngoscope2001;111(8):1479-85.

[4] Couloigner V, Baculard F, El Bakkouri W, et al. Inlay butterflycartilage tympanoplasty in children. Otol Neurotol 2005;26(2):247-51.

[5] Effat KG, Kamal G. Results of inlay cartilage myringoplasty in termsof closure of central tympanic membrane perforations. J Laryngol Otol2005;119(8):611-3.

[6] Anand TS, Kathuria G, Kumar S, et al. Butterfly inlay tympanoplasty:a study in Indian scenario. Indian J Otolaryngol Head Neck Surg2002;54:11-3.

[7] Ghanem MA, Monroy A, Alizadeh FS, et al. Butterfly cartilage graftinlay tympanoplasty for large perforations. Laryngoscope 2006;116:1813-6.

[8] Sakagami M, Yuasa R, Yuasa Y. Simple underlay myringoplasty.J Laryngol Otol 2006;14:1-5.

[9] Rizer FM. Tympanoplasty. Part I: A historical review and a comparisonof techniques. Laryngoscope 1997;107:1-25.

[10] Dornhoffer JL. Hearing results with cartilage tympanoplasty. Laryngo-scope 1997;107:1094-9.

[11] Gerber MJ, Mason JC, Lambert PR. Hearing results after primarycartilage tympanoplasty. Laryngoscope 2000;110:1994-9.