muscle length adaptation, and that its e ects are bilateral. Neurophysiologists, with few exceptions [25], have long believed that version and vergence stimulation, while arising in di erent centers in the brainstem, are com-

2bined into a “final common pathway” at the motoneurons whose axons constitute the motor nerves to the extraocular muscles [26, 27]. In other words, it has been believed that version and vergence stimulation are indistinguishable by the time the impulses reach the extraocular muscles. If that were the case, extraocular muscle length adaptation could not be preferentially responsive to vergence stimulation. Recent evidence suggests, however, that version and vergence signals may indeed remain segregated in the motor nerves and stimulate different fiber types in the extraocular muscles [28, 29]. It is tempting to speculate that those fiber types receiving vergence stimulation are those primarily responsible for muscle length adaptation, but such details have not yet been worked out.

Recent experiments by Joel Miller support the notion of segregation of version and vergence signals by demonstrating that measured extraocular muscle tension shows discrepancies with electrical activity [30]. These observations argue against the final common pathway concept and at least allow the thesis that vergence tonus is primarily responsible for muscle length adaptation.

2.3Changes in Strabismus

However, if the basic muscle lengths change primarily in response to vergence stimulation, how does constant strabismus change over time, when there is presumably no fusional vergence stimulation occurring? It is easy to answer this question in the case of sensory exotropia, because other forms of vergence are occurring.With poor vision in one eye, there is no advantage or incentive to actively align the eyes, or even to converge them when looking up close. With less convergence occurring than before vision was lost in one eye, and at least in older individuals, the normal balance between convergence and divergence is upset in favor of a slight divergence bias, and this divergence bias slowly but actively shortens both lateral rectus muscles and lengthens both medial rectus muscles over time, resulting in increasing exotropia. The deviation, of course, shows up only in the eye with poor vision, until the patient is put under anesthesia, when both the eyes turn out.

Some patients with loss of vision or fusion develop esotropia, especially when vision is lost in early infancy. Vertical misalignment can also develop when vision is lost in one eye. It has been argued before that abnormal

ocular torsion, with associated A and V patterns, are forms of sensory deviations developing over time when fusion is faulty or absent [4].Clearly,the simple decreased need to converge that occurs when vision is lost in one eye cannot explain the development of esotropia, vertical deviations, or torsional deviations. The many di erent ways that strabismus can change over time, if linked to changes in vergence tonus, require a more general explanation.

The explanation, as noted earlier, probably lies in the very nature of biologic control systems. When input to such control systems shuts down, the output rarely goes to zero, but rather goes to a baseline state that may be biased on either side of zero output. In the case of the ocular motor control systems, when the eyes become misaligned enough that fusional vergence cannot operate, retinal image disparities do not result in corrective vergences. In this case, the fusional vergence control mechanisms for horizontal, vertical, and torsional alignment probably do not shut down entirely, but rather decrease their outputs to small nonzero levels, with persistent weak vergence signals biased in one direction or the other, with the direction of this bias depending upon numerous factors.

For example, young children often have a stronger convergence bias than divergence bias, as evidenced by the relative frequency of esotropia vs. exotropia in infancy. This may simply be a manifestation of more hyperopia in childhood, with the attendant increased convergence tonus from accommodative convergence. If vision is lost in one eye in early infancy, it is not surprising that a nonzero convergence bias in the horizontal alignment control system could shorten the medial rectus muscles over time, resulting in sensory esotropia.

Likewise, when fusion is faulty or absent, either primarily or from horizontal misalignment early in life, a baseline output bias in the torsional alignment mechanism can drive the eyes into torsional misalignment with apparent oblique muscle dysfunction and accompanying A and V patterns. The torsion is often seen at first only when awake, disappearing when under anesthesia [31]. Later, as the oblique muscle lengths change, the fundus torsion persists under anesthesia [32]. Still later, after soft tissue remodeling occurs in response to the chronic ocular torsion (the author’s interpretation), the eyes move more along the torted planes defined by the muscle insertions, showing clinical oblique muscle “overaction” (elevation or depression in adduction), and on MRI studies, the connective tissue“pulleys” may be seen to have shifted [33] (the author’s interpretation).

Furthermore, a baseline output bias in the cyclovertical alignment mechanism can drive the eyes into a basic

cyclovertical misalignment, a cyclovertical misalignment which we often call congenital superior oblique paresis, probably mistakenly, because we have no other term for it. Most cases of esotropia are not attributed to sixth nerve palsy, but we persist in attributing many cyclovertical deviations of unknown cause to fourth nerve palsy.

Problems at other points in these control mechanisms can perhaps lead to strabismus in the first place. An abnormality in vergence adaptation has been proposed to cause divergence insu ciency or convergence excess [34]. Poor or absent fusion from birth, in combination with a robust AC/A ratio, could lead to imbalance of muscle length adaptation on the eso side, with progressive esotropia, which we would call congenital esotropia. Alternatively, a higher than normal AC/A ratio [35] could strain fusion su ciently to cause intermittent esotropia, which would then progress to a constant esotropia [2, 3] by the feedback mechanisms just noted. In intermittent exotropia, only a minor defect in fusion could be the initial problem, but as fusion deteriorates, the feedbackdeprived muscle length adaptation mechanism will cause progressive worsening.

Convergence brought into play to damp some forms of nystagmus clearly disrupts the normal alignment control mechanism, leading directly to shortened medial rectus muscles and esotropia. This is the “nystagmus blockage” or “nystagmus compensation” mechanism originally described by Adelstein and Cüppers (cited in [36]). And now that we know that manifest latent nystagmus as well as congenital nystagmus can be damped by convergence [37], this mechanism may be involved in Ciancia’s syndrome as well [38].

2.3.1Diagnostic Occlusion: And the Hazard of Prolonged Occlusion

Diagnostic occlusion of one eye has long been used as a valuable method to break down vergence adaptation to uncover the underlying deviation. Such occlusion will not reverse the e ects of muscle length adaptation in the short term, but will simply reduce the e ects of vergence adaptation over an exponential time course. Thirty to forty-five minutes of monocular occlusion are usually long enough to eliminate most vergence adaptation [13], although diagnostic monocular occlusion for up to 1–2 weeks has been reported.

If diagnostic occlusion is continued for days, eliminating fusion, there is a very real possibility of creating new deviations by the stimulation of new extraocular muscle length adaptation. In the 1920 and 1930s, Marlow advocated occlusion for 7–10 days to fully uncover latent

deviations [39–41]. By careful study of Marlow’s published graphs [40], it is apparent that after 3–5 days of monocular occlusion, significant changes in the monitored deviations often began to appear, and worsen. For example, hyperdeviations and torsional deviations began to appear when there had been none previously. Also, the occluded eye most often developed a hyperdeviation, regardless of which eye was covered, speaking against the uncovering of a latent hyperdeviation [42–44]. Rather than the uncovering of latent deviations, “Marlow occlusion” may indeed have promoted the onset of unguided vergence adaptation and even the onset of muscle length adaptation, with new deviations beginning to occur. The same may be the case in more recent studies by Viirre et al. [45] in monkeys, and by Liesch and Simonsz [46] in normal human subjects. In these studies, new vertical and torsional deviations were noted after 7 days of monocular occlusion of the monkeys and after 3 days of monocular occlusion of the human subjects.

2.3.2Unilateral Changes in Strabismus

Clearly, not all changes in strabismus are bilateral. Patients with loss of fusion from sixth nerve palsy develop an increasingly short and tight ipsilateral medial rectus muscle. The contralateral rectus muscle does not shorten concomitantly. This represents unilateral muscle length adaptation, but from a di erent mechanism. When a skeletal muscle continues to be stimulated but is not stretched out from time to time, it progressively shortens via the active loss of sarcomeres [16]. This is the mechanism demonstrated by Alan Scott by suturing his monkey’s eye temporally [18], and is the mechanism determining changes in the medial and/or lateral rectus muscles in various types of Duane’s syndrome as documented by Collins, Jampolsky, and Howe [47] and by Castañera de Molina and Giñer Muñoz [48].

2.3.2.1Supporting Evidence for Bilateral Feedback Control of Muscle Lengths

What further evidence is there for bilateral feedback control of muscle lengths? We have previously demonstrated that patients with consecutive esotropia following surgery for intermittent exotropia often develop intorsion or extorsion of the eyes, with accompanying oblique muscle overaction and A or V patterns, after having lost fusion for only 1 month [4, 49]. We attribute this to a type of “sensory torsional” deviation due to muscle length adaptation in the torsional dimension.

Weldon Wright, Katie Gotzler, and the author have recently collected a large series of patients with early presbyopia, mostly with deficient or absent fusion, who have developed progressive esotropia probably from the

2increased convergence tonus accompanying the increasing e ort to accommodate. Seeking evidence that such patients are fairly common, we tabulated all the patients that the author had operated on for esotropia over a 17-year period where a reliable onset of the esotropia could be established. Compared with a similar number of patients operated on for exotropia, the esotropia population showed a significantly increased onset of esotropia in their 30s and 40s, as expected [21]. This mechanism, involving muscle length adaptation, is probably responsible for other reports of esotropia developing in adulthood [50, 51] and is similar to the mechanism of hypoaccommodative esotropia occurring in children, as first described by Costenbader [52].

Elizabeth Bell, Adam Bowen, and the author have also identified a series of presbyopic patients, aged 50 years and older, who either had a small amount of uncorrected hyperopia, or who often tried to function without needed correction for near, and developed divergence insu ciency in the later decades of life. They had intermittent or constant esotropia in the distance with diplopia, but could still fuse at near. They are best corrected by bilateral medial rectus muscle recessions [53, 54], with the finding that both medial rectus muscles tend to be tighter than normal by forced ductions at the beginning of surgery. In these patients, we suspect that chronic activation of the near triad [55], which can provide improved visual acuity via slight pupillary constriction, causes increased convergence tonus, leading to shortened medial rectus muscles and the characteristic pattern of divergence insu ciency. Of interest is that the presbyopic patients identified with uncorrected or undercorrected hyperopia showed a somewhat linear increase of distance esotropia with the amount of hyperopia (Bell, Bowen, and Guyton, unpublished).

In the cyclovertical “plane,” which is not really a plane after all, we have long suspected that there should be a thing such as a basic cyclovertical deviation, an analog of straightforward esotropia in the horizontal plane. Recent evidence suggests that the oblique muscles play a much larger role in cyclovertical fusion than previously expected [56–58]. A chronic level of cyclovertical vergence might indeed drive the eyes into a basic cyclovertical deviation, one involving both the vertical rectus muscles and the oblique muscles. But what is this basic cyclovertical deviation? We do not have a name for it. The vast majority of idiopathic cyclovertical deviations are termed congenital superior oblique paresis, or congenital superior oblique palsy. Yet, recent

studies have shown that many patients with these deviations have superior oblique muscles with normal cross-sectional area and normal contractility [59, 60]. Demer et al. wrote in 1995 [59],“Of 19 SO muscles diagnosed to be palsied based on clinical criteria, MRI demonstrated that about half exhibited normal cross-sectional size and contractile characteristics.” Might there be no superior oblique paresis at all in these patients? After all, we do not speak of patients with congenital esotropia as having sixth nerve paresis!

Howard Ying, Nicholas Ramey, and the author are currently investigating the patterns of cyclovertical strabismus that they can create in normal subjects. They have constructed a special haploscope that allows adaptation to increasing vertical, torsional, or horizontal disparities, with near fixation, with fields of view of over 50°, utilizing video-oculography for recording. The entire apparatus can tilt, up to 45°, to the right or left.

To confirm the capability of this apparatus, Fig. 2.7 shows the expected counter roll with head tilt to the right and left before any adaptation.

So far, we have adapted normal subjects to vertical disparities increasing to 6° for 30–45 min. With adaptation, we expect to find that the hyperdeviations induced are accompanied by torsional changes, and that the patterns of misalignment induced, especially with forced head tilting, will help explain the patterns that heretofore have been associated with what is called congenital superior oblique paresis.

The first results appear promising. A normal subject with head straight was slowly adapted over 45 min, maintaining fusion, to an increasing left-over-right

time [s]

Fig. 2.7 Plot of torsional position for each eye shows ocular counter roll with 45° head tilt. A normal subject is continuously recorded with head straight (STR), right head tilt (RHT), and left head tilt (LHT) of 45°. Traces show counter rolling of both the eyes of 4–7°