Abstract
Background:
Achilles tendon lengthening is an important part of the operative management of numerous foot and ankle disorders. The most commonly used percutaneous technique involves three hemitenotomies: medial, lateral, and medial. The discovery that Achilles tendon fibers may undergo external torsion created concern that improper consideration of this feature could result in fiber division insufficient to allow adequate percutaneous lengthening. Varying degrees of torsion have been reported since it was originally described. We hypothesized that torsion of the Achilles tendon does not decrease the effectiveness of a percutaneous lengthening procedure. This hypothesis was tested by comparing the medial-lateral- medial tenotomy pattern to a lateral-medial-lateral pattern. If tendon torsion affects fiber division, then asymmetrical tenotomy patterns should yield different results.
Methods:
Mechanical testing was performed using 22 human fresh tissue lower extremities. Specimens were divided into two groups (11 in each group): “MLM” which underwent testing using a medial-lateral-medial pattern and “LML” which underwent a lateral-medial-lateral pattern. Dorsiflexion load was applied evenly across the forefoot and measured with a force transducer. Dorsiflexion measurements were made from reproducible reference marks. After testing, specimens were dissected to examine tenotomy gap widths and other tissue effects.
Results:
All specimens lengthened successfully. None of the following parameters were statistically different: pre-tenotomy or post-tenotomy dorsiflexion, force to lengthen, total tissue lengthening, and individual gap widths. In no specimen was there complete disruption of the Achilles tendon, and in 18 of 22 specimens (81.8%) the paratenon sleeve was intact except for the tenotomy puncture. In one LML specimen, the sural nerve was partially cut at the distal tenotomy site.
Conclusions:
Regardless of pattern, percutaneous triple-cut Achilles tendon lengthening was effective in achieving additional dorsiflexion at roughly one-third degree per millimeter of tissue lengthening. The pattern of the Achilles tendon lengthening used and, by inference, Achilles tendon torsion had no effect on achievable dorsiflexion. One explanation is that both patterns cut all tendon fibers at least once. We concluded that a properly designed triple-cut tenotomy would be successful regardless of Achilles torsion. Percutaneous Achilles tendon lengthening also appears to be a safe procedure if care is taken to avoid the sural nerve.
INTRODUCTION
Achilles tendon lengthening is an important part of the operative management of numerous foot and ankle disorders. Open procedures for tendon lengthening or recession, as well as many percutaneous techniques have been described. 1 –4,6,9,10,13 One common percutaneous technique, the so-called triple-cut, involves three hemitenotomies through posterior stab wounds in a medial-lateral-medial pattern at 2.5 to 4 cm intervals proximal to the tendon insertion. 12 However, our review of the literature found that the first described triple-cut Achilles tendon lengthening was actually performed anterolateral-posterolateral-anterolateral, all through stab wounds from the medial aspect of the Achilles tendon. 6 This pattern was developed to treat the external torsion of the Achilles tendon as it had recently been described 13 (although percutaneous triple-cut tenotomies had been successfully performed by Hoke well before the first reports of Achilles torsion). 6 The medial- lateral-medial (MLM) pattern is a modification of this original technique, although we could identify no published of its modification.
Torsion, specifically external rotation from proximal to distal, of the human Achilles tendon was first described by White in 1943. 13 In that study, the heel cord is described as undergoing a 90-degree twist. To account for tendon rotation and to ensure that all fibers were divided, White 13 described an Achilles tenotomy in which the medial two-thirds are cut proximally and the anterior two-thirds are cut distally. Subsequent cadaver studies have confirmed that Achilles tendon fibers undergo external rotation, although different degrees have been reported. 2,8,11 Using localizing pins and a protractor, van Gils et al. 11 found a mean torsion of 37 (range 11 to 65) degrees, but no specimen reached 90 degrees. Although they did not quantify the degree of torsion, Cummins et al. 2 performed detailed dissections of 100 specimens and found a wide variation in Achilles torsion, from nearly parallel to complete rotation. The study by Cummins et al. 2 also reported that the double-cut Achilles tendon lengthening described by White 13 left enough lateral fibers intact to theoretically compromise satisfactory lengthening. Similar results were found by Morimoto and Ogata 8 in 264 specimens. None of these studies describes dissection after Achilles tendon lengthening to document the results of their percutaneous procedures.
Several authors have cautioned that disregard of the torsional anatomy of the Achilles tendon during percutaneous Achilles tendon lengthening, or excessive dorsiflexion to lengthen the tenotomized tendon, can lead to tendon rupture. 6,9,11 We found no documentation in the literature to support this assertion, although inadvertent sectioning of the Achilles tendon may well have occurred. Furthermore, double-cut Achilles tendon lengthening has been performed in which three-quarter tenotomies were made only 2 to 3 cm apart and no rupture was reported. 1 Several authors have reported satisfactory outcomes using the White 13 technique or a modification of that technique. 5,10
The purpose of our study was to challenge the assertion that Achilles torsion plays an important role in the safety or efficacy of percutaneous Achilles tendon lengthening by comparing the results of the traditional MLM pattern with a lateral-medial-lateral (LML) pattern. Achilles torsion creates a sagittally asymmetric fiber arrangement. If this torsion affected net fiber division, then a sagittally asymmetric hemitenotomy pattern and its mirror image should yield different results. We hypothesized that external rotation of Achilles tendon fibers plays no detectable role in the results of a percutaneous Achilles tendon lengthening procedure.
Materials and Methods
Twenty-two lower extremity, fresh, human unembalmed tissue specimens were tested. Specimens were either intact cadaver extremities or isolated lower extremities amputated at or above the mid-thigh to preserve the gastrocnemius insertions. These specimens were divided into two groups: 11 underwent the MLM tenotomy pattern (MLM group), and 11 underwent the LML tenotomy pattern (LML group). The original cadaver pool consisted of three males (six specimens) and nine females (18 specimens). Two specimens (both female) were excluded because of other prior dissections rendering them unsuitable for study, leaving 22 intact specimens. Cadavers whose lower extremities were included in the study had a mean weight of 149 lbs and a mean age of 67.1 years at the time of death. Mechanical testing was then performed.
Specimens were positioned prone. Load and angular measurements were performed with the knees flexed and soft tissues prestretched to relieve any postmortem contracture. Dorsiflexion at a tissue-tensioning load (4 kg) and maximal dorsiflexion were then measured. Dorsiflexion load was applied evenly across the metatarsal heads and measured with a Jamar force transducer (Sammons Preston Roylan, Bolingbrook, IL) when quantified. Dorsiflexion measurements were made from reproducible reference lines along the anterior tibial crest and along the lateral glabrous skin border of the foot. The landmark lines were drawn in permanent ink marker before testing to optimize reproducibility, using a long-arm hinged goniometer with 1-degree increments.
Three Achilles tenotomies were then made in an MLM or LML pattern according to group assignment. To simulate our operative technique, all tenotomies were made percutaneously at 3 cm, 6 cm, and 9 cm proximal to the palpated Achilles tendon insertion on the calcaneus. The knee was extended, the ankle dorsiflexed, and the tenotomy sites marked with a permanent ink marker. The Achilles tendon was palpated between the thumb and the forefinger to determine its diameter. A No. 11 scalpel blade was then inserted through the skin and the full thickness of the tendon with the cutting edge facing distally. The scalpel was rotated to face medially or laterally, depending on the placement of the tenotomy. The tenotomy was completed by slowly moving the blade toward the edge of the tendon while applying gentle counter pressure against it. The scalpel was relaxed immediately when the tenotomy was completed by feel. The point of insertion of the blade was just past the midline away from the side of the tenotomy to be made. The intention was to create tenotomies of just greater than 50% of the tendon diameter.
After the tenotomies were performed, dorsiflexion was again measured under a 4-kg tissue-tensioning load. The Achilles tendon was then lengthened by steady, strong dorsiflexion until a palpable and audible pop was heard. Dorsiflexion force at the pop was transcribed. Finally, maximal post-Achilles tendon lengthening dorsiflexion was measured.
After dorsiflexion and load testing, dissection was done to examine the Achilles tendon, its paratenon, and the sural nerve in proximity. Care was taken to preserve these structures after a posterior midsagittal skin incision was made. Once the Achilles paratenon and sural nerves were located, the paratenon was carefully dissected away. The Achilles tendon was inspected and tenotomy gap widths were measured with the knee extended and the ankle in full dorsiflexion. Gap measurements were made at the sagittal midline, using the inset arms of a dial caliper (F.V. Fowler, Boston, MA).
For purposes of this study, safety of the procedures was defined according to two variables: tendon integrity after tenotomy and sural nerve integrity after tenotomy. Three statistical analyses were used to determine significant difference in various data sets. For single variable data differences between the MLM and LML groups, we used the unpaired Student's t-test. For single variable pretreatment and posttreatment data within the MLM or the LML group, and for single variable pretreatment and posttreatment aggregate data (MLM + LML), the paired Student's t-test was used. For individual gap widths, which had three sets per group, analysis of variance (ANOVA) was used.
RESULTS
Using the Student's t-test, none of the measured or calculated parameters between the MLM and the LML group were statistically different with the numbers available (Table 1). Within each group, individual gap widths also were found not to be statistically different (ANOVA p = 0.41 and 0.90 respectively).
Within each group, with the numbers available, the difference between maximal dorsiflexion before and after tenotomy was statistically different (paired Student's t-test p = 0.05 and p = 0.0002, respectively) (Table 2) as was the aggregate difference between maximal dorsiflexion before and after tenotomy (28.1 degrees (SD = 7.83) versus 38.7 degrees (SD = 9.99), paired Student's t-testp < 0.0001). In no specimen was there complete disruption of the Achilles tendon, and in 16 of 22 (72.7%) the paratenon sleeve was intact except for the tenotomy puncture (Figure 1). In one LML specimen (9.0 % of group, 4.5% overall) the sural nerve was partially cut at the proximal lateral tenotomy site.
DISCUSSION
Percutaneous triple-cut Achilles tendon lengthening was effective in achieving additional dorsiflexion at roughly one- third degree per millimeter of tissue lengthening. The pattern of Achilles tendon lengthening and, by inference, Achilles tendon torsion had no effect on achievable dorsiflexion at any stage. This conclusion is further supported by our finding of equal failure forces between groups. This indicates that there is similar tendon fiber integrity regardless of the tenotomy pattern or underlying tendon torsion. As Cummins et al. 2 stated, “the amount of weakening to be obtained can best be judged if the lengthening is taken against the resistance of the tendon.” Assuming full Achilles torsion (90 degrees), each pattern results in three cross-sectional zones where tendon fibers are cut once, three where fibers are cut twice, and a small central zone where fibers are cut by all three tenotomies (Figure 2). Even assuming zero torsion, the slight overlap guarantees that all fibers are cut at least once. There is no zone where fibers are left intact, in contrast to early reports of insufficient lengthening with a double-cut technique. 2,6,8 This is a satisfactory explanation of the similarity between the MLM and LML groups. Coronal plane angulation was not directly measured in this study. Because either pattern cuts all descending fibers at least once, we hypothesize that there would be no significant varus or valgus moment as a result.
Comparisons of biomechanical data between medial-lateral-medial and lateral-medial-lateral groups
TAL = tendo-Achilles lengthening, DF = dorsiflexion, F = force; MLM = medial-lateral-medial; LML = lateral-medial-lateral.; SD = standard deviation.
Dissected specimen demonstrates an intact paratenon sleeve. Only stab punctures are seen at each tenotomy site.
Comparisons of maximal dorsiflexion before and after TAL within each group
MLM = medial-lateral-medial; LML = lateral-medial-lateral.
This technique, with the tenotomy intervals tested, also is safe. In his original article, White 13 stated that the efficacy of the operation could be proven by exposing the lengthened tendon subcutaneously. We agree and, as we were unable to find such an investigation in the literature, included a post-tenotomy dissection in the present study. One study did examine the effect of Achilles tendon lengthening, but the partial tenotomies were done open under direct exposure. 2 Our cadaver dissections revealed that when a cut tendon gives way when stressed (the so-called “pop” of lengthening after the tenotomies are made), its fibers do not stretch or rip but shear past each other in a consistent and controlled manner (Figure 3). No trials resulted in complete tendon disruption, and most preserved the paratenon sheath. Also important, no trial found a failure to lengthen, as has been theorized. It should be noted that the largest tenotomy gap in this study measured 22.1 mm, so performing hemitenotomies closer than 3 cm apart may risk tendon disruption as the cut ends shear past each other. Sural nerve injury did occur once in this study. Care should be taken to control the blade as it emerges from the tendon, particularly on the lateral side, to avoid injury to the sural nerve. This risk, rather than biomechanical results, may be sufficient rationale for preferring a percutaneous Achilles tendon lengthening pattern that minimizes laterally based hemitenotomies.
Schematic diagram shows the net effect of multiple tenotomies assuming 0 or 90 degrees of Achilles tendon rotation. Each pattern of lines represents a single tenotomy. Cross-hatching occurs where the same fibers are cut more than once. Cutting just greater than 50% of the tendon diameter is required to ensure that no central fibers are left intact.
Our study was designed to examine the clinical endpoint of the intervention. Anatomic studies are informative in describing tissue of interest, but are only part of broader investigations if the goal is clinical application of anatomic information. White 13 seems to have provided the impetus for considering tendon torsion when planning a percutaneous Achilles tendon lengthening based on his observation that the current triple-cut technique was “far from satisfactory,” but the nature and extent of the inadequacy of the triple-cut procedure was not discussed. To date, there has been disagreement about the contribution or relevance of Achilles tendon torsion in percutaneous Achilles tendon lengthening. 2 –4,10,11 Review of the literature alone, with the wide variation found in Achilles tendon torsion, suggests that it is incorrect to design a percutaneous Achilles tendon lengthening for a particular degree of torsion. Rather, there should be a single procedure that cuts all tendon fibers with certainty, regardless of the degree of torsion. This study confirms that the MLM pattern (or equally the LML pattern) meets that criterion, with the caveat that the tenotomies must overlap slightly.
While we believe that our study directly investigated whether the Achilles tendon lengthening pattern determines the clinical result, a number of weaknesses do exist that may limit the scope of our conclusions. First, our study design did not randomize specimens into MLM or LML groups; specimens were entered as two serial cohorts. This is not a major deficiency, however, as cadavers were obtained at random with respect to gender, weight, and condition. As the technique involved is simple and established, there should also be little learning curve effect. Second, it would have been possible to include one side of a cadaver to the MLM group and the opposite side of the same cadaver to the LML group. Although this would have added an element of internal control, it also would have contributed to similarity between groups. Since the present study design rejects the null hypothesis with increasing similarity between groups, we do not believe that this is a major deficiency. Third, we were not able to obtain a detailed history on our specimens with respect to the premorbid condition of their tissues. Any previous lower extremity injury or joint contracture was unknown. Increasing the number of enrolled specimens would be the only other method of controlling this bias. Fourth, we used external landmarks rather than reference marks fixed to bone. This does introduce a degree of measurement imprecision, although the surface landmarks used were selected because they were easily identified, constant, and not obscured by habitus or edema in the specimens studied. Fifth, this is a cadaver study. As with many post-mortem studies, one must proceed cautiously when generalizing a conclusion to clinical practice. For example, we concluded that percutaneous Achilles tendon lengthening is a safe procedure because in no case did the Achilles tendon lose continuity. In only one case, one using the LML pattern not currently used in practice, was the sural nerve injured. Because this is a cadaver experiment we can make no claim as to the likelihood of Achilles tendon rupture after a percutaneous Achilles tendon lengthening in vivo, nor can we predict development of a sural neuroma after sural nerve injury. Tendon healing and other patient factors make that claim beyond the scope of this study. Our results do suggest that the Achilles tendon does not get transected during a carefully done percutaneous Achilles tendon lengthening.
Dissected specimen (MLM group) demonstrates a lengthened Achilles tendon. Note the square gaps (result of fiber shear, not stretch) and tendon overlap between gaps. The ankle is in full dorsiflexion.
When performing a percutaneous triple-cut Achilles tendon lengthening, the surgeon's choice of pattern should be influenced by factors other than concern for tissue lengthening and resultant dorsiflexion gain. Consideration of torsional anatomy and placement of hemitenotomies may be more crucial when performing a double-cut Achilles tendon lengthening. However, given the wide and inconsistent variation in Achilles tendon torsion reported in the literature, we recommend the triple-cut Achilles tendon lengthening with each tenotomy cutting slightly more than 50% of tendon diameter. This technique resulted in no failures to lengthen and no tendon ruptures. This pattern is not affected by any degree of Achilles tendon torsion as pattern reversal and cross-sectional analysis both demonstrated. Further investigation may determine if a difference exists between a percutaneous Achilles tendon lengthening using two compared to three hemitenotomies. In addition, we did not specifically examine the coronal plane effect of the Achilles tendon lengthening, if any. Because claims have been made regarding the ability of a percutaneous Achilles tendon lengthening to control a varus hindfoot moment, 3 adding this parameter also would be important for future study.