14.2Clinical and Theoretical Investigations

A series of clinical in vivo investigations of the e ect of di erent methods of handling the SO tendon frenulum, as well as some theoretical calculations made from scale modeling shed important light on how the SO frenulum should be handled when surgery is performed on the SO tendon or superior rectus muscle.

14.2.1The E ect of Superior Rectus Muscle Recession on the Location of the Superior Oblique Tendon Before and After Cutting the Frenulum

This experiment consisted of measuring the posterior displacement of the SO tendon with recession of the superior rectus muscle before and after cutting the SO frenulum in three patients (2, 8, and 25 years of age) who were undergoing enucleation for unrelated reasons [6].At the time of surgery but before the globe was enucleated, the position of the SO tendon was measured before and after cutting the frenulum in the eye undergoing enucleation while suspending the superior rectus muscle at various distances. This was performed as follows: The superior rectus muscle was isolated on a muscle hook, imbricated with two double-armed 6–0 Polyglactin 910 sutures, the check ligaments were cut in the usual manner, and the superior rectus muscle was disinserted. The underlying SO tendon insertion was identified without cutting the frenulum. A single-armed 6–0 Polyglactin 910 suture was sewn into the anterior aspect of the SO tendon midway between the nasal and temporal edge of the superior rectus muscle and knotted in place (Fig. 14.1). A reference knot was tied in this suture approximately 15–20 mm from the knot placed in the SO tendon. Next, with the superior rectus held at the original insertion, the distance between the reference knot and the superior rectus muscle insertion was recorded. This distance was referred to as the initial

Fig. 14.1 Photograph of right eye at surgery as seen from below. The needle of a 6–0 Polyglactin 910 suture is being passed through anterior aspect of the superior oblique (SO) tendon midway between the nasal and temporal edge of the superior rectus muscle with the superior rectus muscle disinserted and reflected upward. The small arrow denotes the SO tendon; the large arrow denotes the reflected superior rectus muscle. (Reprinted from [6] Elsevier Press)

reference knot distance. The superior rectus muscle was then suspended 6, 8, 10, 12, and 14 mm for a total of three recessions at each distance in a randomly generated order to avoid any influence of tissue hysteresis or tissue memory. The temporary suspension of the muscle was accomplished by grasping the sutures in the superior rectus with forceps at the desired distance from the superior rectus and then holding this point on the sutures at the superior rectus insertion. The eye was then rotated to the primary position and the conjunctiva was lifted to verify if the muscle had completely taken up the slack in the suspension suture. If the slack had not been spontaneously taken up for the desired amount of recession, the superior rectus muscle was reposited with instruments and the occurrence thereof noted. The e ect of the superior rectus suspension on the position of the SO tendon was recorded using calipers to measure the distance from the reference knot to the insertion of the superior rectus muscle. This was referred to as the second reference knot distance. A masked assistant (resident, fellow, or scrub nurse) then read the caliper distance using a straight ruler to the nearest 0.5 mm. By subtracting the second reference knot distance from the initial reference knot distance, the amount of posterior movement of the SO tendon was calculated for each successive suspension of the superior rectus muscle (Fig. 14.2).

The SO frenulum was then completely severed under direct visualization by elevating the superior rectus muscle and lysing the connection between it and the underlying SO tendon using sharp and blunt dissection.

All the above measurements were repeated, again with three measurements for each superior rectus suspension distance performed in a randomly determined sequence.

There was essentially a one-to-one correlation between the amount of superior rectus recession and posterior movement of the SO tendon for superior rectus recessions up to 10 mm. After severing the frenulum, there was negligible movement of the SO tendon reaching a maximum of only 1.7 mm in only one patient for a superior rectus recession of 14 mm.

For superior rectus recessions between 10 and 14 mm, the suspended superior rectus typically would not take up the slack to achieve the desired amount of recession prior to severing the frenulum without being manually reposited. This confirms that the frenulum intimately links the superior rectus muscle and the SO tendon. The fact that the superior rectus muscle did not consistently take up the slack for large suspension recessions (10–14 mm) with the frenulum intact, but did so more often when the frenulum was severed, is probably due to a constraining e ect of the frenulum. The frenulum is attached to the SO tendon, which in turn has limited amount of slack to allow the tendon to continue to move freely posteriorly. Hence, at these large recession values, the frenulum may prevent adequate weakening unless the superior rectus muscle is sutured in place. We therefore advocate cutting the frenulum for superior rectus muscle recessions that are larger than 10 mm, especially when using a suspension technique.

In theory, when the frenulum is intact the orientation of the SO tendon would bow backwards as illustrated in Fig. 14.2c when a large recession of the superior rectus muscle is performed. This graphically illustrates why an intact frenulum will limit the amount the superior rectus

muscle can be recessed using a suspension. It appears, however, that this should result in a substantial alteration of the force of the SO muscle. Yet clinically, we do not observe such a profound change in the SO muscle function. One explanation may be that the frenulum allows some movement of the SO tendon relative to the superior rectus muscle during active contraction. Our studies were all done with the patients anesthetized and consequently did not address that possibility.

After cutting the frenulum, the SO muscle moved minimally when the superior rectus muscle was recessed. Because the anterior border of the SO tendon is approximately 8 mm posterior to the superior rectus when the globe is rotated in the downward position, an 8 mm recession of the superior rectus muscle would place its new insertion overlying the SO tendon if the frenulum is severed. The SO insertion is broad and underlies a relatively large area beneath the superior rectus muscle. Consequently, cutting the frenulum may result in di - culty with suturing the superior rectus to the sclera without incorporating some of the SO insertion whose diaphanous nature can make it di cult to visualize. We therefore agree with Jampolsky’s recommendations to preserve the frenulum for superior rectus recessions that are 10 mm or less to insure that the SO tendon will move posteriorly with the recessed superior rectus muscle and not get scarred into the new superior rectus insertion [1, 7]. Furthermore, for recessions greater than 10 mm we advocate lysing this areolar connection owing to its constraining e ect [6].

Although we did not study superior rectus resections [6], we speculate that with the frenulum intact, the SO tendon would be pulled anteriorly with the superior rectus muscle as previously stated by Jampolsky, and the SO

tendon may therefore be at risk of being sutured into the insertion site of the superior rectus muscle [1, 7]. Consequently, for superior rectus resections, we also

advocate separating the frenulum.

14

14.2.2The E ect of the Frenulum

on Superior Oblique Recession Using a Suspension Technique

This experiment consisted of assessing how far the SO tendon retracted (recessed) after disinsertion to simulate what happens with either a recession with a suspension technique or a free disinsertion. This was done both before and after separating the frenulum in a second series of four patients (ages 8, 17, 22, and 47 years) who were undergoing bilateral SO recession using a suspension technique. The position of the SO was measured before and after cutting the frenulum in the following manner: The SO tendon’s insertion was isolated through a superotemporal incision after first hooking the superior rectus muscle. The SO tendon was hooked at its insertion with care to avoid pulling the tendon from under the superior rectus muscle, thus preserving the frenulum. This was done by reflecting the superior rectus nasally as minimally as possible but su cient to allow for visualization of the insertion of the SO tendon. A 6–0 Polyglactin 910 suture was woven through the tendon near the insertion and knotted (Fig. 14.3). A reference knot was tied in the suture 15–20 mm from the distal end of the SO tendon and the superior rectus muscle was set back in its unreflected position. The distance from the reference knot to the temporal edge of the superior rectus muscle

was measured and recorded in the aforementioned masked manner. This was recorded as the initial reference knot distance. The SO tendon was then disinserted, and two successive forced ductions to rotate the eye maximally up and in were performed. With the eye returned to the primary position, the distance between the initial reference knot and the temporal superior rectus edge was remeasured with calipers to give the second reference knot distance. The masked assistant then read the caliper distance using a straight ruler to the nearest 0.5 mm. The amount of recession of the SO tendon was calculated to the nearest 0.5 mm by subtracting the second reference knot distance from the initial reference knot distance. This was repeated for three sets of measurements.

Traction was then placed on the SO tendon, to pull it approximately 12–14mm out from under the superior rectus muscle temporally (Fig. 14.4). This movement essentially brought all of the tendon that is normally under the superior rectus muscle out temporal to it, and e ectively severed the frenulum connection. This maneuver is similar to what frequently occurs if one just exerts substantial traction on the SO tendon when weakening it at the insertion or during a SO tendon tucking procedure. Two forced ductions were again performed to rotate the eye up and in. The distance between the knot and the superior rectus edge was measured with calipers in the same manner as when the frenulum was intact. Again, using simple subtraction, the amount of recession of the SO tendon after the frenulum was stripped was calculated using our masked measurement technique for three successive measurements.

To control the possibility that the amount of recession simply increased with the multiple forced ductions that were needed to obtain multiple measurements, a single set of

Fig. 14.4 Surgical photograph of the right eye rotated downward as viewed from below; superior muscles are at the top in the photograph. Traction is placed on the SO tendon pulling it 12–14 mm out from under the superior rectus muscle temporally to e ectively sever the frenulum. Small arrow denotes SO tendon; large arrow denotes suture tied to the cut end of the SO tendon. (Reprinted from [6] Elsevier Press)

measurements was taken prior to and after stripping the frenulum on the other (control) eye in the same manner as in the first (study) eye. In two patients, the study procedure was performed in the right eye first, and in the other two patients, the study procedure was performed in left eye first.

The mean distance that the SO tendon recessed was 2.4 ± 0.4 mm before cutting the frenulum and 8.5

± 0.7 mm after cutting the frenulum. There was a statistically significant di erence between the two measurements (P = 0.0011, paired two-tailed student’s t-test). The same procedure was followed in the fellow control eye for one set of measurement. For the control eyes the mean recession prior to stripping the frenulum was 2.4

± 0.3 mm and after stripping the frenulum was 8.0

± 0.8 mm (P = 0.0004, paired two-tailed student’s t-test). These values for the amount of recession obtained in the control eyes before and after stripping the frenulum were essentially identical to the values for the study eyes, despite the control eyes only having a single measurement. This confirms that taking multiple measurements prior to stripping the frenulum was not a confounding factor on the amount that the SO moved after stripping the frenulum.

The results of this experiment are consistent with the observation that the maximal e ect of a recession of the SO tendon using a suspension technique can only be

achieved by cutting the frenulum [4]. It also suggests that asymmetric e ects may occur with bilateral SO recession using a suspension technique, if there is asymmetric stripping of the frenulum. On the other hand, stripping the frenulum may allow the disinserted SO tendon to migrate forward resulting in the SO tendon incarceration syndrome [2]. Thus how the frenulum is handled with these procedures may be a matter of tradeo s.

14.2.3The Theoretical E ect of the Superior Oblique Frenulum on the Posterior Partial Tenectomy of the

Superior Oblique

The threefold function of the SO muscle includes intorsion, depression, and to a lesser degree, abduction. These actions are uniquely related to its anatomy and the angle the tendon makes with the anterior–posterior axis. The SO tendon makes an angle of approximately 54° with the anterior–posterior axis. The anterior fibers of the SO tendon make a relatively large angle with the anterior–poste- rior axis and therefore are thought to primarily have a torsional action, and only a small vertical action. Prieto Diaz calculated the relative vertical and torsional actions of the anterior and posterior fibers of the SO tendon using computer-aided design software and determined that the vertical action is approximately 1/3 of the torsional action [8]. The posterior fibers of the SO tendon make a smaller angle with the anterior–posterior axis than the anterior fibers. He concluded they therefore contribute approximately 50% less torsion than the anterior fibers but twice as much vertical action.

These anatomic considerations of the differential effects of the anterior and posterior fibers of the SO tendon have given rise to different surgical procedures depending on whether one wants more torsion vs. vertical correction. For example, the Harada–Ito operation tightens the anterior fibers and primarily provides torsional changes [9]. Conversely, the posterior partial tenectomy primarily weakens the more posterior fibers of the SO tendon and thus gives more vertical correction with minimal change in torsion. Prieto–Diaz first described this procedure, which consists of cutting the posterior 4/5 or 7/8 of the SO tendon at its insertion and then excising a posterior triangle of tendon extending about 8–12 mm toward the trochlea [10, 11]. He proposed this operation to surgically treat A-patterns without affecting torsion. It has been reported to be effective in collapsing A patterns of up to 20 PD; however, it is not effective in decreasing the overdepression in adduction resulting

in a pseudo-SOOA [3, 5] (Fig. 14.5). Why this procedure fails to address the overdepression in adduction has not been adequately explained. We feel that some unique considerations about the SO frenulum as well as some anatomic considerations of the SO tendon explain why the posterior partial tenectomy operation does not eliminate the overdepression in adduction.

To study this, we used scale figures of the anatomy of the SO and SR obtained from Orbit™1.8 (Eidactics, San Francisco, CA) to determine the angles made by the anterior and posterior fibers of the SO tendon with the ante- rior–posterior axis when the eye was in the primary position, as well as in adduction. We then modified those figures to assume that the frenulum constrained the SO tendon to the SR muscle and recalculated the same angles. The contribution of the net force directed parallel to the anterior–posterior axis represents the force that creates depression, and the contribution of the net force directed perpendicular to the anterior–posterior axis represents the torsional force. The percentage of original SO force that is directed vertically and torsionally is the cosine and sine of the angle made by the SO tendon and the ante- rior–posterior axis, respectively, multiplied by 100.

Figure 14.6a shows the eye in primary position. The anterior fibers of the SO tendon make an angle of 75° with the anterior–posterior axis. Thus, the torsional force vector of these fibers is the sine of 75°, or 0.97 times the magnitude of the net force. Or in other words, the torsional force vector equals 97% of the net force. Similarly, the vertical force vector is the cosine of 75° multiplied by 100, or 26% of the net force.

When the eye is adducted 35°, and if one assumes the frenulum constrains the tendon to the SR muscle, the tendon will bow backwards as shown in Fig. 14.6b. In this picture, which is modified from the Orbit™1.8 model, we

have kept the distance between the anterior edge of the SO tendon and the SR insertion the same, implying that the constraining property of the frenulum completely prevents the SO tendon from slipping anteriorly. In this scenario, the original angle made by the anterior fibers of the SO tendon and the anterior–posterior axis is approximately the same. As seen in Fig. 14.6b, the anterior fibers of the SO tendon still make an angle of 75° with the ante- rior–posterior axis. Consequently, in the normal nonoperated eye, the contribution of the SO forces of intorsion, abduction, and depression remain relatively unchanged in adduction compared with the primary position.

Figure 14.6c illustrates the situation after a posterior partial tenectomy procedure. The excised portion of the posterior four fifths of the SO tendon insertion is outlined in black. This surgical procedure necessitates that the frenulum be excised, which allows the SO tendon to move forward. This substantially decreases the angle between the anterior fibers of the SO tendon and the anterior–posterior axis. In Fig. 14.6c, we measured this angle to be approximately 40°. In this position, the depressor action of the SO tendon is increased compared with that found in the unoperated state. The magnitude of depression is the sine of 40° or 77% of the total net force as compared with only 26% prior to the surgical procedure. This may be one explanation why overdepression in adduction persist after posterior partial tenectomy. This residual abnormality of versions may be due to the unavoidable excision of the SO frenulum, which occurs with this surgical procedure, and the e ect this has on the subsequent distribution of vertical force of the SO tendon. Persistent overdepression in adduction has been reported as occurring in 40.4% [12]–57% [5] of patients after posterior partial SO tenectomy. Despite this unwanted overdepression in adduction, weakening of the

SO with posterior partial tenectomy e ectively reduces the exo-shift in down gaze and thus reduces the A pattern [5, 10–12]. This may be due to the ability of the adducting power of the inferior rectus muscle to prevail over any residual abducting power of the weakened SO in the adducted and depressed position (unpublished written personal communication from A. Castanera de Molina, July 18, 2007). However, overdepression occurs even when the A-pattern is e ectively collapsed, suggesting that this motility pattern is not simply due to a surgical undercorrection. Castanera considers this common postoperative complication of downshoot in adduction to be a direct consequence of the surgery itself (unpublished written personal communication from A Castanera de Molina, July 18, 2007). This would be consistent with our hypothesis that excision of the frenulum can result in forward slippage of the remaining fibers of the SO when the eye is adducted, thus increasing their vertical force.

Some investigators have speculated that the downshoot in adduction seen after partial posterior SO tenectomy occurs secondary to a limitation of depression in abduction of the contralateral eye after bilateral surgery.

This results in a pseudo-SOOA in the ipsilateral eye by Herring’s law [5, 8]. There are several theories as to the cause of this limitation. For example, anteriorization of the SO tendon insertion to a preequatorial location after a posterior partial tenectomy has been theorized. Using the Orbit™ 1.8 model, Castanera simulated that an anterior shift of the muscle insertion centroid of 4.45 mm after a posterior partial tenectomy would cause a reduction in the vertical force of the SO tendon [13]. He also modeled the situation in which the cut end of the SO tendon could inadvertently be reattached to the sclera, thus simulating a recession plus resection procedure. Both simulations show a similar change in the vertical force component such that the SO tendon becomes an elevator in abduction with no change of depression in adduction. Another cause of the limitation to depression in abduction of the contralateral eye may due to iatrogenic incarceration of the SO tendon to the SR insertion [2, 5, 13]. This complication also places the e ective insertion of the SO tendon to a preequatorial position. One further mechanism could be the presence of underlying occult SR contracture [7]. We feel that