Outcomes following ossicular chain reconstruction with composite prostheses: hydroxyapatite-polyethylene vs. hydroxyapatite-titanium.
We conducted a retrospective study to compare the results of ossicular chain reconstruction (OCR) with two types of composite prosthesis: a hydroxyapatite-polyethylene (HAPEX) implant and a hydroxyapatite- titanium (HATi) prosthesis. We reviewed the records of 222 patients--104 males and 118 females, aged 8 to 79 years (mean: 39.7)--who had undergone OCR for ossicular chain dysfunction and who met our eligibility criteria. In addition to demographic data and the type of prosthesis, we compiled information on pre- and postoperative audiometric findings, the underlying diagnosis, the timing of surgery (primary, planned, or revision), the type of surgery (tympanoplasty alone, tympanoplasty with antrotomy, intact-canal-wall tympanomastoidectomy, or canal-wall-down tympanomastoidectomy), the extent of reconstruction (partial or total), the use of the malleus, the use of a tragal cartilage graft, and evidence of extrusion. Of the 222 patients, 46 had undergone insertion of either a partial (n = 36) or total (n = 10) ossicular replacement prosthesis (PORP and TORP, respectively) made with HAPEX, and 176 had received a PORP (n = 101) or TORP (n = 75) made with HATi. Postoperatively, the mean air-bone gap (ABG) was 14.0 dB in the HAPEX group and 14.7 dB in the HATi group, which was not a significant difference (p = 0.61). Postoperative success (ABG [less than or equal to] 20 dB) with PORP was obtained in 30 of the 36 patients in the HAPEX group (83.3%) and in 87 of the 101 patients in the HATi group (86.1%), while success with TORP was achieved in 7 of 10 HAPEX patients (70.0%) and 56 of 75 HATi patients (74.7%); there was no significant difference in either PORP or TORP success rates between the HAPEX and HATi groups (p = 0.32). A significantly better hearing result was obtained when the malleus was used in reconstruction (p = 0.035), but the use of tragal cartilage led to a significantly worse outcome (p = 0.026). Revision surgery was associated with a significantly worse postoperative result (p = 0.034). Prosthesis extrusion was observed in 9.0% of all cases. The two types of composite assessed in this study yielded similar results in terms of functional hearing and stability, but the HATi prosthesis had some significant advantages. For example, it was associated with more cases in which the ABG closed to less than 10 dB. In addition, because of its thinner stem and lower profile, it can be used in situations that are not possible with the HAPEX implant.
The primary goal of the surgical treatment of chronic otitis media is the eradication of disease; a secondary goal is to correct a conductive hearing loss. Achieving the second goal often requires reconstruction of the ossicular chain. The use of autologous materials, if available, is a compatible reconstruction option. However, in diseased ears, the availability of autologous materials such as the incus for interposition grafting is limited. Therefore, a wide variety of alternatives have been developed.
The ideal material for ossicular chain reconstruction (OCR) is one that (1) is safe, (2) is associated with only limited adverse tissue reactions, (3) can be integrated into normal surrounding tissue, (4) is easy to manipulate during surgery, (5) provides a good hearing result, and (6) is stable over long-term follow-up. To overcome the limitations of some biomaterials, composite prostheses have been developed. These prostheses offer biocompatibility at the interface with the tympanic membrane and a shaft that is easily modified during surgery.
[FIGURE 1 OMITTED]
The HAPEX Universal Prosthesis (Gyrus ENT; Bartlett, Tenn.) features a hydroxyapatite-polyethylene composite (HAPEX) shaft that can be easily modified (figure 1, A). The distal shaft is narrow so that it can fit on the footplate and function as a total prosthesis. If needed, the device can be trimmed to create a wider shaft for use as a partial prosthesis that fits over the capitulum of the stapes.
The Grace Alto Titanium + Hydroxyapatite prosthesis (Grace Medical; Memphis, Tenn.) is a lightweight, adjustable-length implant that is available in both total and partial ossiculoplasty configurations (figure 1, B). It is smaller and lighter than the HAPEX prosthesis.
Both prostheses have hydroxyapatite heads that can be placed directly under the tympanic membrane. In addition, an offset groove or tab on the head allows each to be secured under the malleus.
In this article, we describe our comparison of hearing outcomes and complication rates among patients who received either of these two prostheses.
Patients and methods
We reviewed the records of 256 consecutive patients who had undergone OCR at the Michigan Ear Institute between Jan. 2, 2000, and July 31, 2008. For a case to be included in our study, a patient's chart had to contain complete pre- and postoperative audiometric data, operative notes, and follow-up information. Of the 256 patients, there were 222-104 males and 118 females, aged 8 to 79 years (mean: 39.7)--who satisfied these eligibility criteria.
In addition to demographic data and the type of prosthesis, we compiled information on pre- and postoperative audiometric findings, the underlying diagnosis, the timing of surgery (primary, planned, or revision), the type of surgery (tympanoplasty alone, tympanoplasty with antrotomy, intact-canal-wall tympanomastoidectomy, or canal-wall-down tympanomastoidectomy), the extent of reconstruction (partial or total), the use of the malleus, the use of a tragal cartilage graft, and evidence of extrusion.
Audiometry. In accordance with the 1995 guidelines of the Academy of Otolaryngology-Head and Neck Surgery's Committee on Hearing and Equilibrium guidelines, (1) we used the four-frequency (500, 1,000, 2,000, and 3,000 Hz) air-bone gap (ABG) obtained at least 4 months after surgery as the primary outcome measure. OCR success was defined as a postoperative ABG of 20 dB or less.
Underlying diagnosis. More than half the patients (57%) had chronic otitis media with cholesteatoma as the cause of their ossicular chain dysfunction, and 30% had chronic otitis media without cholesteatoma. Among the rest, 10% presented with congenital cholesteatoma, 2% with trauma, and 1% with lateral chain fixation.
Surgical procedures. All patients had been operated on by the senior author (D.I.B.). Of the entire study population, 129 patients underwent tympanoplasty alone (58.1%), 12 patients underwent tympanoplasty with antrotomy (5.4%), 66 patients underwent intact-canal-wall tympanomastoidectomy (29.7%), and 15 patients underwent canal-wall-down tympanomastoidectomy (6.8%). Overall, 51 patients underwent primary surgery (23.0%), 86 had planned surgery (38.7%), and 85 had revision surgery (38.3%). The distribution in the HAPEX group was 13, 14, and 19, respectively, and the distribution in the hydroxyapatite-titanium (HATi) group was 38, 72, and 66.
[FIGURE 2 OMITTED]
The malleus was used in OCR whenever possible; 168 patients received a malleus (75.7%) and 54 did not (24.3%).
Tragal cartilage was used in 35 patients (15.8%) who had severe atelectasis of the tympanic membrane and evidence of poor eustachian tube function, leaving 187 patients who did not receive cartilage (84.2%). The tragal cartilage was shaved to half-thickness and placed over the prosthesis head. An over-under graft was used for tympanic membrane grafting, as previously described by Kartush et al. (2)
The HAPEX prosthesis was sized by placing it within the middle ear and trimming it with a surgical scalpel. The HATi prosthesis was cut to the appropriate size by using the length gauge included in the packaging. A small nipper was used to trim the shaft of the prosthesis. Further refinement of the length was accomplished by adjusting the titanium shaft, either longer or shorter, in increments of 0.2 mm per click.
A total of 113 patients had surgery on the right ear (50.9%), and 109 patients had surgery on the left (49.1%).
Type of prosthesis. We divided the 222 patients into two groups on the basis of the type of prosthesis they received:
HAPEX group. The HAPEX group was made up of 46 patients aged 11 to 77 years (mean: 36.3). Thirty-six of them (78.3%) received a partial ossicular replacement prosthesis (PORP), and the other 10 (21.7%) received a total prosthesis (TORP).
HATi group. The HATi group was made up of 176 patients aged 8 to 79 years (mean: 36.2). A PORP was inserted in 101 of them (57.4%) and a TORP in 75 (42.6%).
Follow-up. The duration of postoperative follow-up ranged from 10 to 82 months (mean: 42.9) in the HAPEX group and from 6 to 53 months (mean: 28.9) in the HATi group. The reason for the longer follow-up in the HAPEX group was that the senior author had used the polyethylene composite for several years before switching exclusively to the titanium composite.
The average interval from surgery to audiometry was 21.4 months in the HAPEX group and 16.3 months in the HATi group. All audiograms were obtained at least 4 months after surgery.
Statistical analysis. Statistical analysis was performed with the Statistical Package for the Social Sciences software (v. 15.0 for Windows; SPSS; Chicago). Data were analyzed using paired t tests, repeated measures ANOVA, one-way ANOVA, and Pearson chi-square tests.
HAPEX vs. HATi. In the HAPEX group, the mean ABG fell from 33.2 dB preoperatively to 14.0 dB ([+ or -] 5.7) postoperatively. In the HATi group, the corresponding values were 34.0 and 14.7 dB ([+ or -] 8.2). The difference between postoperative ABGs in the two groups was not statistically significant (p = 0.61).
A successful outcome was achieved in 37 of 46 patients in the HAPEX group (80.4%) and in 143 of 176 in the HATi group (81.3%); again, the difference was not statistically significant (figure 2). Although the difference was not significant, the HATi prosthesis was associated with a greater percentage of cases in which the ABG closed to 10 dB or less (34.7 to 26.1%).
Timing of surgery. Patients who underwent revision surgery had a significantly worse outcome than did those who underwent primary surgery or planned surgery (p = 0.034). There were no other significant differences among these three groups.
Type of surgery. The only significant difference relating to the type of surgery was seen in the HATi patients who underwent the canal-wall-down procedure; their mean postoperative ABG was significantly worse than that of all other subgroups (p = 0.08).
[FIGURE 3 OMITTED]
Extent of reconstruction. Postoperative success (ABG [less than or equal to] 20 dB) with PORP was obtained in 30 of the 36 patients (83.3%) in the HAPEX group and in 87 of the 101 patients (86.1%) in the HATi group. Success with TORP was achieved in 7 of the 10 patients (70.0%) in the HAPEX group and in 56 of the 75 patients (74.7%) in the HATi group. There was no significant difference in either PORP success rates or in TORP success rates between the HAPEX and HATi groups (p = 0.32).
Within the HAPEX group, the mean ABG fell to 13.8 dB in the 36 patients who received a PORP and to 14.7 dB in those who received a TORE The difference in postoperative ABG between PORP and TORP surgery in the HAPEX group was not statistically significant (p = 0.667) (figure 3).
Within the HATi group, the mean ABGs fell to 13.4 dB in the PORP subgroup and to 16.3 dB in the TORP subgroup. In this case, the difference between the two postoperative values was statistically significant (p = 0.021) in favor of PORP (figure 3).
Malleus vs. no malleus. The use of the malleus in O CR was associated with a significantly greater improvement in hearing (figure 4). Overall, the ABG met the benchmark for success ([less than or equal to] 20 dB) in 146 of the 168 patients who received a malleus (86.9%) and in 39 of the 54 patients who did not (72.2%). The difference between the two groups was statistically significant (p = 0.035) in favor of malleus use.
Cartilage vs. no cartilage. Among the entire study population, the use of a shaved tragal cartilage graft over the prosthesis was associated with a significantly worse postoperative ABG (p = 0.026) (figure 5).
Within the HAPEX group, the mean postoperative ABGs in the cartilage and no-cartilage subgroups were 14.7 and 13.9 dB, respectively--not a significant difference (p = 0.763). Within the much larger HATi group, however, the difference was significant (p = 0.013); the corresponding values were 18.0 and 13.9 dB, respectively.
[FIGURE 6 OMITTED]
Extrusion. Complications were uncommon, as extrusion occurred in only 20 patients (9.0%). All extrusions were associated with atelectasis of the tympanic membrane.
There were 3 extrusions (6.5%) in the HAPEX group. In 2 of these cases, the prosthesis remained under the malleus. Two of the extrusions occurred late in the course of follow-up. Cartilage was not used in any of these cases.
Of the 17 extrusions (9.7%) in the HATi group, the prosthesis remained under the malleus in good position in 6 patients; these 6 patients had intact hearing, and they were followed conservatively (figure 6, A). A tragal cartilage graft was used in 3 of these cases, and the malleus was used in 10.
In this study, the hearing results obtained with the HATi prosthesis were comparable to those obtained with the HAPEX implant, but the HATi device had some significant advantages. For example, it was associated with a greater percentage of cases in which the ABG closed to 10 dB or less (34.7 to 26.1%). In addition, because of its thinner stem and lower profile, it can be used in situations that are not possible with the HAPEX implant.
In our series, we used the malleus in reconstruction of the ossicular chain in 168 of the 222 patients (75.7%). Hearing outcomes were significantly better when the malleus was used in reconstruction. The presence of the malleus handle has been reported to be an important element in the stabilization of the prosthesis and for the improvement in sound transmission. (3-5)
The use of shaved tragal cartilage over the prosthesis head was associated with a lower success rate than was OCR without a cartilage overlay graft. Coupling the prosthesis head with the tympanic membrane is of utmost importance to the transfer of sound energy to the stapes superstructure or stapes footplate. While the inertial load of a piece of attached cartilage, which can add as much as 30 mg of mass, has not been shown to have a great effect on the transfer function of an ossicular prosthesis, our series showed that patients who did not have cartilage placed over the hydroxyapatite head had significantly better results.
Both types of prosthesis in our study had a hydroxyapatite head to allow for placement directly under the tympanic membrane. Both types of implant are easy to manipulate during surgery, and both can be trimmed to fit rapidly. In addition, the HATi prosthesis provides the surgeon with the ability to make small changes in implant length in the event that it was not trimmed perfectly the first time. The ideal prosthesis design provides a tight, permanent coupling between the tympanic membrane and the stapes, and small alterations in length (on the order of 0.2 to 0.4 mm) can have a significant impact on the final ABG. (6) The ability to make small adjustments in prosthesis length after the initial trimming allows for optimal placement.
Total ossicular reconstruction was performed in a greater percentage of the HATi group (75 of 176 [42.6%]) than the HAPEX group (10 of 46 [21.7%]). The HATi TORP can be used in patients whose anatomy would preclude the use of a HAPEX or other nontitanium implant, owing to the HATi's extremely narrow shaft and visibility when placing the prosthesis (figure 6, B).
A partial prosthesis placement can result in less than ideal outcomes if the angle of the stapes superstructure is poor (i.e., leaning superiorly or inferiorly), if there is a fracture or partial weakening of the stapes, if the capitulum is eroded, or if the malleus is in a very anterior position. In such cases, we often use the total titanium prosthesis because it results in a better ossicular chain architecture.
Studies have suggested that the angle formed by the prosthesis in the tympanic membrane should be less than 30. In selected cases, bypassing the stapes superstructure and using a TORP may better afford this.
Previous studies of ossicular reconstruction have found extrusion rates with hydroxyapatite prostheses ranging from 4 to 21.6%. (6-9) The use of cartilage to reinforce the tympanic membrane has been shown to decrease the rate of prosthesis extrusion. (10) In our series, we did not see any association between extrusion and the use of cartilage in reconstruction. Extrusion in our patients was universally associated with atelectasis of the tympanic membrane, signifying that poor eustachian tube function was a contributing factor.
While there were 17 extrusions in our 176 HATi patients (9.7%), 6 of these "extruding" prostheses were positioned under the malleus, and there was only minimal exposure of the hydroxyapatite head noted during routine postoperative follow-up. However, for statistical purposes, these cases were still considered to represent an extrusion. No intervention was required for these patients, and their hearing remained intact. These patients were followed routinely, and they did not require any further surgery. Therefore, the incidence of true extrusion that required revision surgery was only 6.3% (11 of 176), which was nearly identical to the incidence of 6.5% (3 of 46) in the HATi group. Placing the prosthesis under the malleus stabilizes the implant and may decrease the likelihood that severe retraction of the tympanic membrane will lead to prosthesis extrusion.
In conclusion, we report hearing outcomes in a large series of OCRs performed by a single surgeon using composite prostheses. Postoperative hearing results were comparable in the HAPEX and HATi groups, but there were significant advantages associated with use of the HATi prosthesis. The use of the malleus in OCR was associated with better hearing results, and the use of tragal cartilage was associated with worse results. The hydroxyapatite head on both of these prostheses facilitates placement without the need for a cartilage interposition graft. Complication rates were similar with the two prostheses.
Both the HAPEX and HATi prostheses offer ease of use, and both are associated with excellent hearing outcomes and low complication rates. The HATi prosthesis, because of its thinner stem and low profile, offers significant advantages because it can be used in situations in which it is not possible to use the HAPEX implant.
(1.) American Academy of Otolaryngology-Head and Neck Surgery Foundation. Committee on Hearing and Equilibrium guidelines for the evaluation of results of treatment of conductive hearing loss. Otolaryngol Head Neck Surg 1995;113(3):186-7.
(2.) Kartush JM, Michaelides EM, Becvarovski Z, LaRouere MJ. Overunder tympanoplasty. Laryngoscope 2002;112(5):802-7.
(3.) Albu S, Babighian G, Trabalzini E Prognostic factors in tympanoplasty. Am J Otol 1998;19(2):136-40.
(4.) Goldenberg RA, Emmet JR. Current use of implants in middle ear surgery. Otol Neurotol2001;22(2):145-52.
(5.) Pasha R, Hill SL III, Burgio DL. Evaluation of hydroxyapatite ossicular chain prostheses. Otolaryngol Head Neck Surg 2000;123(4):425-9.
(6.) Zhao S, Hato N, Goode RL. Experimental study of an adjustable-length prosthesis in a temporal bone model. Acta Otolaryngol2005; 125(1):33-7.
(7.) Truy E, Naiman AN, Pavillon C, et al. Hydroxyapatite versus titanium ossiculoplasty. Otol Neurotol2007;28(4):492-8.
(8.) Smyth GD. Five-year report on partial ossicular replacement prostheses and total ossicular replacement prostheses. Otolaryngol Head Neck Surg 1982;90(3 Pt 1):343-6.
(9.) Stupp CH, Dalchow C, Grun D, et al. Three years of experience with titanium implants in the middle ear [in German]. Laryngorhinootologie 1999;78(6):299-303.
(10.) Kobayashi T, Gyo K, Shinohara T, Yanagihara N. Ossicular reconstruction using hydroxyapatite prostheses with interposed cartilage. Am J Otolaryngol2002;23(4):222-7.
Yoav Hahn, MD; Dennis I. Bojrab, MD
From the Dallas Ear Institute (Dr. Hahn) and the Michigan Ear Institute, Farmington Hills (Dr. Bojrab). The study described in this article was conducted at the Michigan Ear Institute.
Corresponding author: Yoav Hahn, MD, Dallas Ear Institute, 7777 Forest Lane, A-103, Dallas, TX 75230. Email: email@example.com
Figure 4. Chart shows the percentage of patients with a successful outcome according to malleus use. Malleus use was associated with a significantly better outcome (p = 0.026). All patients Malleus used (n = 168) 86.9% No malleus used (n = 54) 72.2% Note: Table made from bar graph. Figure 5. Chart shows the percentage of patients with a successful outcome according to use of a cartilage graft. Use of a cartilage graft was associated with a significantly worse outcome All patients Cartilage used (n = 35) 68.6% No cartilage used (n = 187) 85.6% Note: Table made from bar graph.