Two-stage procedures (Nagata1–5 and Firmin6) for reconstruction of microtia have recently been increasingly used. In the first stage, the lobule is transposed, and the complete auricle framework, including the tragus, is inserted into a skin pocket. In the second stage—the elevation of the framework—an additional cartilage block (or blocks) is placed behind the framework for distinct projection. The axial flap of temporoparietal fascia with skin graft is widely used to cover the added cartilage block(s). A temporoparietal fascia flap is reliable with its rich vascularity, but its use is associated with some morbidity, including long scars on the temporal region and alopecia. Moreover, harvesting of a temporoparietal fascia flap deprives the surgeon of the option for significant traumatic and secondary reconstruction.
Given these concerns, minimum invasion of the temporoparietal fascia is required for refinement of the procedure. Firmin6 and Brent7,8 have proposed a postauricular turnover fascial flap to wrap the added cartilage block. Furthermore, in the context of preserving an axial temporoparietal fascia flap, the authors devised a technically simplified covering procedure by creating a pocket in the temporoparietal fascia, and named the procedure the “temporoparietal fascia pocket method.”
PATIENTS AND METHODS
Microtia patients who do not wish (or have an indication) for middle ear surgery are candidates for the method described here. We routinely collaborate with otorhinolaryngology surgeons in the second stage of reconstruction and perform both middle ear surgery and framework elevation. The middle ear surgery needs elevation of a temporoparietal fascia flap.
In addition, it is preferable that microtia patients do not have other combined severe hypoplastic disease, such as Treacher Collins syndrome, because the temporoparietal fascia pocket method requires sufficient temporoparietal fascia volume. However, we can still perform the temporoparietal fascia method by switching to an axial temporoparietal fascia in cases where the temporoparietal fascia volume is inadequate for the pocket method flap.
The surgical technique is outlined in Figure 1. An incision is made peripheral to the border of the helix, and the constructed auricle is elevated from the head, dissecting superficially to the temporoparietal fascia for the posterior margin of the concha. Then, an incision is made in the temporoparietal fascia and the capsule to create a slit perpendicular to the temporal head plane. A pocket is created with dissection between the capsule and the cartilage laterally, and between the temporoparietal fascia and the pericranium medially. The pillars of rib cartilage are inserted into the pocket and fixed to the base of the auricular cartilage and to the pericranium on the temporal bone with absorbable sutures. After the slit is closed, the temporoparietal fascia and the capsule on the cartilage are covered with a 0.010- to 0.016-inch split-thickness skin graft harvested from the temporal scalp. The sutures on the skin graft are left long so as to be tied over a bolster to tamponade the graft to the recipient bed. Before skin grafting, the retroauricular skin is advanced toward the ear to reduce the skin graft requirement.
Estimation of Projection Angles and Statistical Analyses
The angles of the temporoauricular sulci were estimated by calculating inverse trigonometric functions of the preoperative and postoperative widths of the auricles on profile images. The landmarks and the lines in the profile imageswere defined as follows: point paratragion (pT) is the superiormost point of the tragus or the clearly identifiable point around it when the contours of the tragus are obscure; line M is the line from paratragion to exocanthion; and point paranasion (pN) is the intersecting point on line M and the curve of the dorsum of the nose. For comparisons among images of different sizes, the unit distance was measured as the length of the long axis of the reconstructed auricle if it was not clearly distorted, or the distance between two identifiable points such as nevi or two ends of a hairpin. The width of the auricle (defined as W) was measured as the distance from paratragion to the intersecting point on line M and the posterior curve of the helix of the auricle. The angle of the temporoauricular sulcus was then calculated as cos-1(Wpre/Wpost), where Wpre and Wpost are the preoperative and postoperative widths of the auricle, respectively (Fig. 2).
Because profile photographs taken in daily outpatient follow-up are not standardized, like a cephalogram, parallax distortion because of shifting from the true profile photograph must be managed. For management of the differences in several angles of cranial rotation on the cephalocaudal axis, measurements of the distance from paratragion to paranasion were obtained, and the difference in the cranial rotation angle between two photographs taken at a distinct time point in the same patient was estimated. The angle differences were calculated by the approximation formula that is shown as Δ in Figure 3, which graphically demonstrates the cranial rotation in a top-down view, where α and β are the angles from the anterior part of the midsagittal plane and are positive values when they rotate clockwise. In addition, α and β are relatively small angles, because side view-intended photographs do not deviate markedly from the true profile (i.e., 0 degree). A cranial shape is approximated as an ellipse with a ratio of the major semiaxis to the minor semiaxis of 1:0.8, adopting the mesocephalic cranial index.
The accuracy of the approximation formula was evaluated using three-dimensional computed tomographic head images of 10 individuals without congenital craniofacial conditions at several angles in profile rotated in 5-degree increments from −20 to 20 degrees, and measurements were obtained for angle difference estimation. The one-sample t test was used for statistical comparisons between the actual differences and calculated differences in each rotation angle combination of α and β.
The mean discrepancies and 95 percent confidence intervals are shown in Table 1. The difference was not significant (df = 9, p > 0.05, one-sample t test) in many combinations of α and β. Even in other combinations of α and β where the difference was significant, the gaps from the actual angle differences of rotation were small. Therefore, our rotation difference estimation formula is acceptable for practical use, especially in small angle combinations of α and β.
Collectively, the angles of the temporoauricular sulci, the projection angles, were calculated by the change in width of the auricles, adding the correction angles to address parallax distortion. Even though there is the risk of deviation from the true angle, the accuracy of our evaluation was much better than subjective evaluation, such as “excellent,” “good,” and “fair.” In addition, the estimation system can evaluate the projection of the auricle by the photographs, no matter how old they are. To evaluate the long-term stability of the method, two-way analysis of variance (p < 0.05) using IBM SPSS Version 23 software (IBM Corp., Armonk, N.Y.) was carried out to analyze the influence on the projection angles of the method (an axial temporoparietal fascia flap method versus a temporoparietal fascia pocket method) over short-term and long-term follow-up.
A total of 38 reconstructed ears in 38 patients with microtia ranging in age from 9 to 19 years were elevated using the authors’ method from 2002 to 2014 and followed for at least 5 months. The average age at surgery was 11 years. The patients included 31 male patients and seven female patients. The mean follow-up period was 47 months.
For estimation of projection angles and statistical analyses, 27 auricles of 26 patients elevated with use of the standard axial temporoparietal fascia flap method, 25 auricles elevated along with middle ear surgery, and 27 auricles of 27 patients elevated with the temporoparietal fascia pocket method without complications that had appropriate follow-up photographs for measurement were enrolled. The projections of the auricle of both methods were assessed in short-term and long-term follow-up (Table 2). The projection of the auricle was retained in long-term follow-up: 23 ± 8 degrees with the standard temporoparietal fascia method and 28 ± 11 degrees with the temporoparietal fascia pocket method, which is considered to be comparable to the average cephaloauricular angles reported by da Silva Freitas et al.9 as 31 ± 6 degrees. They defined the cephaloauricular angle as the intersection of a straight line running through the tragus insertion and the lateral portion of the mastoid region with a straight line running through the tragus and the middle of the helix. Because the present measurement point of the helix is posterior to theirs, the present calculation of the cephaloauricular angle is smaller by several degrees than theirs.
Two-way analysis of variance demonstrated that the elevation method and follow-up time did not significantly affect the projection angle of the auricle (method, p = 0.06; time, p = 0.46; method × time, p = 0.55; two-way analysis of variance). Accordingly, the temporoparietal fascia pocket method appears to have long-term stability and efficacy equal to the standard temporoparietal fascia flap method. Furthermore, taking into account that the long-term follow-up time of the temporoparietal fascia pocket method was longer than that of the standard temporoparietal fascia flap method (df = 52, p = 0.03, unpaired t test), it may be concluded that our method has stability superior to the standard temporoparietal fascia flap. Furthermore, the mean difference of the projection angle between the axial temporoparietal fascia flap method and the temporoparietal fascia pocket method was −4.4 degrees (95 percent CI, −9.1 to 0.23 degrees, Bonferroni post hoc test). Thus, despite the difference not being significant, there could be a tendency that the projection angle of the temporoparietal fascia pocket method is larger than that of the axial temporoparietal fascia flap method. This tendency can be explained by the position where the pillars of the rib cartilage are placed. In cases of elevation with middle ear surgery, the pillars are placed posterior to the reconstructed auditory meatus. Therefore, the projection angle of elevation with middle ear surgery is smaller than that of elevation without middle ear surgery, if the height of the cartilage pillars is the same.
Collectively, good projection of the auricles was achieved after elevation by the temporoparietal fascia pocket method in cases with no complications. Furthermore, the projection of the auricles showed little tendency to become effaced over long-term observation, and it showed equal efficacy to the standard temporoparietal fascia flap method.
The overall complication rate related to ear reconstruction, including infection and skin graft failure, was 13.2 percent. Cartilage exposure caused by infection was observed in one case (2.6 percent). Local débridement of the framework, a part of the helix, was needed in that case, and the resultant auricular deformity was later modified. Partial skin graft failure was observed in three cases (7.9 percent), but the ulcers healed with conservative management. However, in all three of the cases, contracture resulted in shallow sulci, which needed further operations for depth; in each of the cases, a postauricular flap, an axial temporoparietal fascia flap with skin graft, and a V-Y advancement flap of the superior end of the sulcus were used, respectively. The infection rate and skin graft necrosis rate were higher than the rates reported by Long et al.,10 who reviewed 60 articles on autologous cartilage microtia reconstruction, as 0.9 percent and 0.41 percent, respectively, which may not fully mirror the complication rate. The cartilage pillars collapsed in one case (2.6 percent) and were corrected to an upright position and covered with a postauricular flap.
The patient in case 1 was a 10-year-old boy with left lobule-type microtia (Fig. 4). Six months after cartilage framework grafting, the ear was elevated with the temporoparietal fascia pocket method. The six pillars of costal cartilage were bundled together and inserted into the temporoparietal fascia pocket. The retroauricular skin was advanced toward the ear to reduce the skin graft requirement, and the sulcus was covered with a 0.010-inch split-thickness graft harvested from the temporal region. The reconstructed sulcus was well defined 1 year postoperatively.
The patient in case 2 was a 10-year-old boy with right lobule-type microtia (Fig. 5). Five months after cartilage framework grafting, the ear was elevated with the temporoparietal fascia pocket method. The four pillars of costal cartilage were bundled together and inserted into the temporoparietal fascia pocket. The retroauricular skin was advanced toward the ear to reduce the skin graft requirement, and the sulcus was covered with a 0.012-inch split-thickness graft harvested from the temporal region. The reconstructed sulcus was well defined, even in the relatively long period after ear elevation at 4 years 10 months.
The authors devised a less invasive method for cartilage covering by creating a pocket in the postauricular temporoparietal fascia. The temporoparietal fascia pocket method requires a minimal incision in the temporoparietal fascia and no additional incision in the temporal region. The method is straightforward, and excellent results in ear reconstruction have been achieved with it.
Auricular separation solely with skin grafting leads to gradual postoperative fading of projection. To counteract the considerably strong shrinking force of the skin graft, an additional cartilage block is placed behind the framework for distinct and durable projection. The added bare cartilage block must be covered with tissue to provide nourishment for skin grafting. A temporoparietal fascia flap with its rich vascularity is reliable. Thus, Nagata5 uses the fascia as an axial flap to cover a cartilage block.
An axial temporoparietal fascia flap, however, is also an appropriate option for significant traumatic and secondary reconstruction cases. Therefore, sparing an axial temporoparietal fascia is desirable. Firmin6 and Brent7,8 proposed a retroauricular turnover fascial flap to wrap the added cartilage block.
We addressed this issue by decreasing the invasion to the temporoparietal fascia, creating a slit and a pocket in the temporoparietal fascia. Our method is simple, but it does not impair the vascularity of the temporoparietal fascia, and it keeps the fascia in the anatomically original place. Firmin and Marchac11 reported a tunnel technique that adds moderate projection to the framework. They create a tunnel behind the framework to bury a piece of cartilage under the retroauricular soft tissue. This tunnel can be dissected behind the antihelix to achieve maximal projection of the upper portion of the ear, behind the antitragus to project the lobule, or occasionally behind both. Because they did not describe the details of the tunnel technique and their results, it appears that they prefer a modified Nagata technique to achieve appropriate projection. In contrast, our technique can mobilize the framework to achieve appropriate projection and fix the cartilage to the framework and the temporal region. Although it can be speculated that their tunnel has a common space with our pocket, our dissected extent must be larger than their tunnel. Moreover, our incision is located at the center of the space. Therefore, we think a pocket is an appropriate name for the created space. Walton and Beahm12 suggested that Brent performed the following in the third stage of his microtia reconstruction repairs: “The ear position is stabilized by placing a piece of banked costal cartilage posteriorly beneath the framework in a ‘fascial’ pocket.” In this context, a fascial pocket means a turned-over occipitalis fascia flap from behind the ear, with a long curved fascial incision posteriorly. Brent7 covers a scalp-banked rib cartilage graft with a retroauricular turnover fascial flap. Beheiry and Abdel-Hamid13 reported that “The [temporoparietal fascia] was found to be part of the subcutaneous fascial system, being mobile and continuous in all directions with other structures of that system: the galea superiorly, the frontalis muscle anteriorly, the occipitalis posteriorly, and the auricularis muscles and the superficial musculoaponeurotic system of the face inferiorly.” In addition, Park et al.14 reported that “On the upper portion of the retroauricular surface, between the skin and the temporal bone, there are three discrete fascial layers: the superficial temporal fascia, the deep temporal fascia, and between them the innominate fascia.” Regarding the lower portion of the retroauricular surface, they described the superficial mastoid fascia and the deep mastoid fascia that could be elevated from the underlying thin fascia investing the sternocleidomastoid musculoaponeurotic portion. The superficial mastoid fascia corresponds to the superficial temporal fascia cephalad, and the deep mastoid fascia corresponds to the innominate fascia. We minimally incise the temporoparietal fascia directly under the reconstructed auricle and create a pocket. Because the superficial temporal artery, the posterior auricular artery, and the occipital artery remain intact, it is feasible to use an axial temporoparietal fascia flap, a mastoid fascia flap, and an occipital fascia flap after our approach.
Moreover, our method has the advantage of ear projection durability and shows efficacy equal to the standard axial temporoparietal fascia flap. The temporoauricular sulci have a tendency to hold their steep profile over a long period. Projection durability also demonstrated that the pillars of the cartilage are less likely to be resorbed, which in turn suggests that the cartilage of the framework is less likely to be resorbed, because the framework is close to the skin and has the same or richer vascularity than the pillars of cartilage.
In addition, even though the thickness of skin graft was in the realm of ultrathin (0.010- to 0.016-inch), the present study suggested that the thinness of the skin graft does not affect the projection durability. However, shrinkage of the skin graft always occurs, which is seemingly inconsistent with projection durability. To address this issue, there are two possible explanations: contraction of the temporoparietal fascia is less likely because of less damage to vascularity; and the shrinking force of the skin graft is directed anteriorly to push the pillars of the rib cartilage orderly encapsulated by a nearly intact temporoparietal fascia, thereby sustaining the projection of the auricle as a buffering function.
Although our method has technical advantages of being simple and less invasive, there are occasions when sufficient volume of the temporoparietal fascia is not available to create an appropriate pocket. These circumstances could arise as a result of an associated facial deformity, such as hemifacial microsomia, Goldenhar syndrome, and Treacher Collins syndrome, in each of which the temporal muscle is often hypoplastic. In such cases, we switch to a procedure with an axial temporoparietal fascia flap.
Even though projection of the auricles was sustained in the long term, the form of the auricle changed consistently after surgery, as is the case in any other reconstruction method for microtia. Remodeling and resorption of cartilage as a result of limited repair capabilities play some role in decay of the auricle. However, in view of the fact that cartilage demands less oxygen than skin, contraction of skin appears to be the main force collapsing the auricle.
Our system for estimation of the projection angle has a risk of deviation from the true projection angle by several degrees. When an auricle before elevation surgery is tilted along the line that extends anterolateral to posteromedial (medially projected), especially in a patient with a small cranium, the projection angle can be underestimated. However, the ability of humans to recognize parallax is not so poor. It contributes to risk reduction by discarding inappropriate photographs that are apparently deviated from the profile view and have medially projected auricles. Furthermore, besides rotation on the axial plane, rotation on the coronal plane can occur, although parallax on the coronal plane affects the length for the observer minimally because it is multiplied by the cosine function, which is nearly 1 when the rotation angle is small.
In addition, growth of the face may affect the measurement. However, because cranial growth reaches 96 percent of adult size at the age of 10 years (the age group that commonly undergoes auricular reconstruction), there is a negligibly small change in the distance from paratragion to paranasion after surgery.
Despite the limitations in its use, the technical advantages of our temporoparietal fascia pocket method (i.e., it is simple and less invasive) are clear. Furthermore, our method creates excellent, long-lasting projection of the reconstructed ear. We think that the temporoparietal fascia pocket method can be a standard choice for elevation of a reconstructed auricle.
The temporoparietal fascia pocket method is simple and maintains distinct projection of the constructed auricles for a long period. Moreover, it is less invasive and has the benefit of sparing temporoparietal fascia flap elevation.