Figure 1. Three vectors are marked on the patient, originating at the lateral orbital rim and inserting on the nasolabial fold at the nasal ala, midfold, and oral commissure. A, The lateral orbital rim is exposed and stab incisions are made at each nasolabial fold site. B, A long Keith needle is used to pass a suture subcutaneously from the lateral orbital rim to each stab incision. C, Each suture is then passed through a small polytetrafluoroethylene pledget to prevent suture loop migration. D, The sutures are passed back to the lateral rim, and an appropriate degree of tension is applied to each. E, Improved oral commissure position and redefined nasolabial fold are present at completion of the procedure.
Figure 2. This patient underwent static facial suspension following extirpation of a left parotid adenoid cystic carcinoma requiring resection of the facial nerve. A, In the early postoperative period (1 month after surgery), the overcorrection seems excessive and the patient was initially dissatisfied with his appearance. B, After suture stretch and elongation, the patient has an acceptable final result at 21 months after surgery.
Figure 3. The Mini Bionix testing apparatus (MTS Systems Corporation, Eden Prairie, Minn) with suture loop in place and knot positioned on the upper disk to roughly simulate positioning in vivo on the lateral orbital rim.
Figure 4. Sutures are placed along 3 vectors in our static facial suspension technique. A, Pledgets are shown at 3 sites along the nasolabial fold: the nasal ala, midfold, and oral commissure. There is a right angle created by lines drawn from the lateral rim to the nasal ala and along the nasolabial fold. Potential force distribution can be calculated using right triangles created by vectors passing through the nasolabial fold. B, If a total load of 70 N is present, this load could be distributed as shown along vectors representing each of the 3 sutures. This depicts performance that could be expected using size 3-0 polypropylene suture (Prolene; Johnson & Johnson, New Brunswick, NJ). C, A total load of 150 N could be borne with load sharing when using size 0 polybutilate-coated braided polyester (Ethibond Excel; Johnson & Johnson) or polyester impregnated with polytetrafluoroethylene (Tevdek; Teleflex, Inc, Limerick, Pa). Such performance is comparable to that of a single piece of human acellular dermis (AlloDerm; LifeCell Corporation, Branchburg, NJ) or expanded polytetrafluoroethylene (Gore-Tex; W. L. Gore & Associates, Flagstaff, Ariz).
Humphrey CD, McIff TE, Sykes KJ, Tsue TT, Kriet JD. Suture Biomechanics and Static Facial Suspension. Arch Facial Plast Surg. 2007;9(3):188-193. doi:10.1001/archfaci.9.3.188
Author Affiliations: Departments of Otolaryngology–Head and Neck Surgery (Drs Humphrey, Tsue, and Kriet and Mr Sykes) and Orthopedic Surgery (Dr McIff), The University of Kansas Medical Center, Kansas City.
Correspondence: Clinton D. Humphrey, MD, Department of Otolaryngology–Head and Neck Surgery, The University of Kansas Medical Center, Kansas City, KS 66160 (email@example.com).
Background Static facial suspension (SFS) continues to play a role for rehabilitation in patients with facial paralysis. We perform SFS almost exclusively with a suture technique in our practice. Monofilament polypropylene suture (Prolene) is commonly used for SFS, but we have witnessed occasional failure and some stretching with this material. The purpose of this study was to establish and compare the biomechanical properties of 3 suture types—polypropylene, polybutilate-coated braided polyester (PBCP) (Ethibond Excel), and braided polyester impregnated with polytetrafluoroethylene (PIP) (Tevdek)—to assess their suitability for SFS.
Methods Six samples of 0, 2-0, and 3-0 polypropylene, PBCP, and PIP were tested. The mean load to failure was calculated for each suture type. Stiffness and elongation at specific loads were calculated to compare stretch between materials.
Results The load to failure of PBCP and PIP was significantly greater than that for polypropylene for all suture sizes. In addition, PBCP and PIP had significantly less elongation than did polypropylene at clinically relevant loads.
Conclusions Both PBCP and PIP had superior load-bearing properties and decreased stretch when compared with polypropylene. These properties suggest that, for SFS with suture, use of PBCP or PIP may reduce the incidence of breakage and elongation, improving outcomes.
Static facial suspension (SFS) continues to play an important role in the rehabilitation of facial paralysis in select patients. The procedure continues to evolve because new materials and simpler, more efficient techniques have been used with improved results. Fascia lata, expanded polytetrafluoroethylene (PTFE) (Gore-Tex; W. L. Gore & Associates, Flagstaff, Ariz), and human acellular dermis (AlloDerm; LifeCell Corporation, Branchburg, NJ) are among the numerous materials that have been used for SFS in the past. Although they are effective for suspension, human acellular dermis and expanded PTFE require extensive dissection for positioning in the midface. The dissection increases operative time and makes revisions or reversal difficult at best. In addition, our experience with expanded PTFE has been marred by frequent wound complications, which have been observed by other authors as well.1-2 To avoid some of these disadvantages, a trend has developed in our own practice toward the use of SFS with suture.
We have adopted a technique similar to that described by Alex and Nguyen.3 Three vectors are marked on the patient that originate at the lateral orbital rim and insert on the nasolabial fold at the nasal ala, midfold, and oral commissure regions. The lateral orbital rim is exposed and stab incisions are made at each nasolabial fold site. A long Keith needle is used to pass a suture subcutaneously from the lateral orbital rim to each stab incision. Each suture is then passed through a small PTFE pledget to prevent suture loop slippage and migration. The sutures are passed back to the lateral rim, and an appropriate amount of tension is placed on each to recreate the nasolabial fold and maintain a horizontal configuration of the oral aperture. A bone tunnel is created in the lateral orbital rim through which the suspension sutures are secured (Figure 1).
Advantages of this suture technique vs alternative techniques and materials include minimal dissection, shorter operative time, ease of revision, and a decreased number of complications. The procedure is also easily reversed and can be performed more feasibly as a delayed procedure if necessary. Similar advantages to performing SFS with suture have been cited by other investigators.3 Although many suture options exist, we believe that familiarity has led most reconstructive surgeons to use predominantly polypropylene suture (Prolene; Johnson & Johnson, New Brunswick, NJ) for this procedure. We too initially adopted polypropylene but have observed some disadvantages to the use of that material. Size 3-0 polypropylene sutures have been observed to fail during the early postoperative period.2 Patients report a snap or popping sound when the facial suspension fails, suggesting that suture breakage has occurred. A second disadvantage is the need to anticipate and estimate inevitable suture stretch or relaxation. This necessitates initial surgical overcorrection with all 3 sutures, but especially at the oral commissure, to obtain a satisfactory long-term result (Figure 2). Breakage and elongation are major disadvantages of the suture technique when using polypropylene and, prior to this study, we had no experience using alternative suture materials. Complications due to breakage and elongation are possible but less frequent when using expanded PTFE or human acellular dermis for SFS. Although the biomechanical properties of expanded PTFE and human acellular dermis as used in SFS have been studied, suture materials when used in this capacity have not.4-5 A comparison of the biomechanical properties of different suture types for use in SFS may provide guidance in selecting optimal suspension materials that promote improved patient outcomes. Manufacturers perform biomechanical testing on their suture materials during research and development, but such testing is not performed to emulate the context of SFS and we have found the data difficult to obtain. Our aim in this study was to perform objective biomechanical testing on various suture types to assess their suitability for use in SFS.
Six samples each of 0, 2-0, and 3-0 polypropylene, polybutilate-coated polyester (PBCP) (Ethibond Excel; Johnson & Johnson), and polyester impregnated with PTFE (PIP) (Tevdek; Teleflex,Inc, Limerick, Pa) were tested to failure using testing conditions similar to those in a previous study by our group of expanded PTFE and human acellular dermis.2 A materials testing machine (Mini Bionix; MTS Systems Corporation, Eden Prairie, Minn) was used to apply tension to single loops of each suture tied with a single knot consisting of 8 square knots. The suture loops were tied around two 13-mm-diameter polished brass disks, which, when moved apart by the testing machine, applied tension to the loops. The knot was always positioned on the upper disk to roughly simulate the position where it normally sits on the lateral orbital rim (Figure 3). The gauge length of suture between disk axes was 44 mm. Based on measurement of an SFS suture explanted from a patient after failure, actual gauge length in vivo can be as little as half that value. This smaller length is consistent with the 22-mm gauge length used in the previous study of expanded PTFE and human acellular dermis.2 Doubling the gauge length in the present study was necessary to allow adequate space for the materials testing machine to function properly with the brass disks and suture loops in place. Changing the gauge length does not alter load to failure data; stiffness and elongation change in inverse proportion to sample length and can be adjusted easily for cross-comparison between studies. The testing machine was used to increase axial loading at a rate of 20 mm/min until each sample failed.
The failure site of suture samples was always near the knot but never on the disks, suggesting that there were no variables introduced by the disks (eg, irregular surface or friction) that induced suture failure. This is consistent with findings in failed SFS sutures explanted from patients that demonstrated breakage near the knot (unpublished observations). The mean maximum load to failure was calculated for each suture type.
Stiffness was calculated as the slope of the curve when load was plotted as a function of elongation. This curve was linear for monofilament polypropylene samples, and stiffness could be calculated. Polybutilate-coated polyester and PIP are polyfilament braided sutures; these yielded curves that were reproducible but not linear. Therefore, slope was always calculated for PBCP and PIP by using the initial segment of the curve, where maximum stiffness was present. In addition, we calculated elongation in millimeters at specific load values of 20 and 30 N as an alternative method for comparing the amount of stretch that occurred among the groups having linear and nonlinear stiffness. This calculation allowed us to quantify the amount of overcorrection required for each suture during surgery to compensate for stretching that occurs at loads likely to be encountered clinically. Means were obtained and statistical analysis was performed using the analysis of variance between groups and Bonferroni-Dunn tests (α = .05).
The results for mean load to failure are shown in Table 1. The load to failure for polypropylene was significantly lower than that for PBCP or PIP in all suture sizes. The load to failure for PBCP vs PIP was never statistically different. As expected, differences in load to failure were always statistically significant between suture sizes regardless of the material tested (eg, 0 suture made of any material had a higher mean load to failure than all types of 2-0 or 3-0 suture). All suture samples of a given material and size failed within a small range of load values, which is typical of synthetic materials (Table 2).
Polypropylene demonstrated significantly less stiffness than either PBCP or PIP in all suture sizes (Table 3). Size 2-0 PIP demonstrated significantly greater stiffness than 2-0 PBCP; however, the stiffness of PIP and PBCP was not statistically different in the 0 and 3-0 samples. Elongation at loads of 20 and 30 N was significantly greater for polypropylene than for PIP or PBCP in all suture sizes. Size 3-0 PIP had significantly greater elongation than PBCP at a load of 20 N; however, elongation of sizes 0 and 2-0 PBCP and PIP was not statistically different at loads of either 20 or 30 N.
Load-bearing ability and elongation with applied force are properties of critical importance to functional outcome in SFS. A lower load to failure increases the likelihood of suture breakage during the early postoperative period. We have clinically observed early breakage of 3-0 polypropylene suture on several occasions. Breaks have occurred most often along the oral commissure vector. The mean maximum load to failure for 3-0 polypropylene, the suture we had previously chosen most frequently for SFS, was approximately 30 N. The alternative suture materials PBCP and PIP, which were evaluated in this study, tolerated significantly larger loads before failure in all suture sizes. If selected for SFS, these materials hypothetically would be less likely to fail in the early postoperative period.
Based on the load to failure of 3-0 polypropylene in this study and our clinical experience with failure at the oral commissure when using this material, one may conclude that it is not uncommon to place a load of 30 N along this vector in vivo. Elongation data were collected at 20 and 30 N for this reason; we believe that these are loads frequently encountered clinically on individual sutures in SFS. Substantial elongation of the suture material will compromise functional and aesthetic outcomes. The elongation of size 3-0 polypropylene suture was greater than 25% (or >1 cm) when a force of 30 N was applied. Elongation of size 3-0 PBCP and PIP was around 10%. Size 0 PBCP and PIP elongate only about 5%. The shorter gauge length used in vivo would decrease the amplitude of elongation witnessed clinically, but these results unequivocally demonstrate that polypropylene suture loaded with 20 to 30 N will consistently elongate 2 to 3 times more than either PBCP or PIP. Likewise, stiffness values were significantly greater for PBCP and PIP than for polypropylene. In a study of sutures used in tendon repair, Lawrence and Davis6 also concluded that polyfilament braided polyester sutures such as PBCP provide greater stiffness than polypropylene did. Static facial suspension may not maintain oral competence and facial appearance when there is inadequate stiffness and significant stretch occurs. Increased early elongation seems to forecast a propensity for greater creep over time as well, but this study cannot definitively prove such a claim.
Previous studies by Morgan et al4 and Sclafani et al5 have examined the biomechanical properties of human acellular dermis (AlloDerm) and expanded PTFE (Gore-Tex) as used in SFS. Data obtained by Morgan et al for the load to failure and stiffness of these materials is shown in Table 4. Stiffness is inversely proportional to sample gauge length, and stiffness data from the earlier study4 were appropriately halved in this table to compensate for the longer gauge length used in the present study. These stiffness values were then used to calculate the elongation that would occur in these materials with a load of 30 N. The mean load to failure of either of these materials initially appears to be much greater than that for any of the suture groups. However, placing 3 sutures allows for load sharing along 3 vectors. Load sharing would be equal if the sutures were placed in parallel, essentially tripling the load to failure, but, because the sutures have vectors acting at different angles, the load is not evenly distributed. This is determined in part by the surgeon, who must select the appropriate amount of tension for each suture before tying and securing it at the lateral rim. In our experience, more tension is required along the oral commissure vector to prevent oral ptosis.
Figure 4A demonstrates the suture placement and vectors used in the SFS technique. There is an approximately 90° angle between lines drawn from the lateral orbital rim to the nasal ala and along the nasolabial fold. This right angle can be used to calculate the force acting along any vector originating from the rim and passing through the nasolabial fold as the hypotenuse of a right triangle. Using this supposition, potential force distribution can be calculated between 3 sutures placed at the aforementioned sites along the nasolabial fold in SFS as in Figure 4A. In Figure 4B and C, the larger blue arrow represents a single vector along which a human acellular dermis or expanded PTFE sling might act. This same load is distributed along 3 vectors at different angles when a suture sling is used. We considered the clinical load that we believe is placed on each suture in vivo to estimate distribution between and to calculate the load placed on the 3 smaller vectors, represented by the red arrows. The sum of the x and y components of these 3 smaller vectors is equal to the x and y components of the larger vector, but the sum of the 3 smaller vectors themselves does not equal that of the larger vector because of the differences in angles at which multivector (eg, suture) and single-vector (eg, human acellular dermis or expanded PTFE) slings act (ie, the sum of 17, 30, and 26 is 73, not 70). The loads in Figure 4B represent performance that could reasonably be expected when using size 3-0 polypropylene suture, based on our data. The total load borne by the 3 sutures is greater than what would be expected from a single suture but still does not compare favorably with human acellular dermis and expanded PTFE. Figure 4C demonstrates the hypothetical load sharing when using size 0 PBCP or PIP.
The load to failure of both human acellular dermis and expanded PTFE in the previous study by Morgan et al4 exceeded 100 N. Size 0 PBCP or PIP could offer similar performance, with total load-bearing capacity exceeding 150 N. A limitation of any comparison between suture and materials such as human acellular dermis and expanded PTFE is that, unlike suture, different widths of other materials can be tested and wider samples may increase load-bearing capacity. Morgan et al used 15-mm-wide samples of human acellular dermis and expanded PTFE. The sheets used for SFS are typically wider than 15 mm, and the testing of sheets equaling the average width used clinically might be more useful for comparison with suture. The gauge length in the earlier study was 22 mm, which we compensated for as described in the “Methods” section. Once stiffness is adjusted for this gauge length differential, human acellular dermis has stiffness similar to sizes 0 and 2-0 PBCP and PIP. Expanded PTFE is less stiff and its performance is more similar to that of polypropylene.
We tested no biological materials in our study, but the study by Morgan et al emphasized the differences in load to failure between synthetic and biological materials. Table 2 demonstrates that human acellular dermis, a biological material, failed across a broad range of load values. Conversely, expanded PTFE, a synthetic material, failed across a narrow range.4 The synthetic suture materials we tested also performed very consistently, failing over a narrow range. A study by Choe et al7 also demonstrated this increased variability with human acellular dermis. The inconsistent performance of human acellular dermis occurs because the dermis must be harvested, which results in irregularities and weaker areas throughout most pieces. Expanded PTFE provides consistent performance but has been associated with frequent wound complications and extrusion in irradiated patients. In our hands, suture provides the consistent performance of a synthetic material without the wound complications seen with expanded PTFE.
Late postoperative failure of SFS has more varied causes than does early postoperative failure. Breakage may still occur but is probably more related to permanent structural alterations in a material over time than to initial load-to-failure capacity. A definite limitation to our study is that load to failure does not estimate how suture will perform over time, although this is an area of ongoing study in our laboratory. Late failure may also occur because there is sawing of distal tissues subjected to the continual suspension force. A PTFE pledget is used to minimize this problem; however, we have witnessed 2 cases in which both the pledget and the suture loop have remained intact but still dissected superiorly through the subcutaneous tissue. Biomechanical suture properties have no bearing on this process, and revision procedures have been required in these cases. Creep may also result in late failure of SFS with suture. A material's initial stiffness and elongation may or may not predict the propensity for and degree of creep that will occur over time.
Knowledge of the load to failure, stiffness, and elongation characteristics of different materials may assist the facial plastic surgeon in predicting the likelihood of SFS success or failure. Polypropylene suture performed poorly in all of these areas compared with PBCP and PIP. Many biomechanical properties of polypropylene are not ideal for SFS with suture. Higher load-bearing capacity could reduce the incidence of suture breakage. Decreased need for initial overcorrection to compensate for suture elongation and stretch would allow the surgeon to produce more predictable and consistent results. The biomechanics of PBCP and PIP sutures also compare favorably with human acellular dermis and expanded PTFE.
The differences between PBCP and PIP, as they relate to properties that will affect SFS performance, seem to be minimal, and either material seems to offer an excellent alternative to polypropylene for this application. A small number of patients in our practice have already undergone SFS using PBCP with good results. An added benefit has been that the braided PBCP is more malleable and less palpable over the lateral orbital rim than polypropylene is. We propose the adoption of heavy polyfilament braided-polyester sutures such as PBCP or PIP when a static suspension procedure is indicated for rehabilitation of patients with facial paralysis.
Accepted for Publication: January 11, 2007.
Author Contributions:Study concept and design: Humphrey, Tsue, and Kriet. Acquisition of data: Humphrey and McIff. Analysis and interpretation of data: Humphrey, McIff, Sykes, and Kriet. Drafting of the manuscript: Humphrey, Sykes, Tsue, and Kriet. Critical revision of the manuscript for important intellectual content: Humphrey, McIff, Tsue, and Kriet. Statistical analysis: Humphrey and Sykes. Administrative, technical, and material support: McIff. Study supervision: McIff, Tsue, and Kriet.
Financial Disclosure: None reported.