- Email: [email protected]
Ocular plagiocephaly: ocular torticollis with skull and facial asymmetry1
Ocular Plagiocephaly: Ocular Torticollis with Skull and Facial Asymmetry Marc F. Greenberg, MD, Zane F. Pollard, MD Purpose: To observe facial asymmetry in patients with ocular torticollis to better understand its cause. Design: Observational case series. Participants: Forty-four consecutive patients with ocular torticollis in one author’s (MFG) private practice were examined for facial asymmetry from January 1998 to August 1998. Some of these, as well as selective others before January 1998 were photographed for a total of 53 photos. Methods: One author (MFG) examined the frontal, maxillary, and mandibular facial areas of 44 consecutive torticollis patients for appearance of unilateral compression or reduced mass. The laterality of such findings was compared with the side of the head turn or tilt. Photographs of 53 selected ocular torticollis patients were inspected and the direction of nasal tip and columella deviation compared with the direction of head tilt or turn. Main Outcome Measures: Subjective clinical determination of appearance of unilateral facial compression or reduced facial mass. Subjective photographic determination of nasal tip and columella deviation. Results: Forty-three patients with 10 types of ocular torticollis examined were included. Forty-one of 43 showed compression or reduced mass on the same side as the head turn or tilt, including 3 with adult-onset strabismus. Eight nonsuperior oblique palsy patients had nasal deviation to the same side as the torticollis. Six of the eight had head tilts. Seventeen nonsuperior oblique palsy patients had nasal tip deviation to the opposite side of the torticollis. All were pure head turns. Eight superior oblique palsy patients had nasal tip deviation to the same side as the torticollis; nine had deviation opposite. Conclusions: Patients with multiple types of ocular torticollis, including face turns, show similar appearance of facial compression on the side of the torticollis, suggesting that the tilt or turn itself may cause the asymmetry. This includes face turn strabismus, in which facial asymmetry has not previously been described. Head tilts are frequently associated with nasal tip deviation to the side of the torticollis, head turns with deviation opposite. We refer to such asymmetric facial changes associated with ocular torticollis as “ocular plagiocephaly.” Ophthalmology 2000;107:173–179 © 2000 by the American Academy of Ophthalmology. As early as 1958, Parks1 described facial asymmetry associated with torticollis from congenital cyclovertical extraocular muscle palsy. More recently, Wilson and Hoxie2 reported seven of nine congenital superior oblique palsy patients with “a shallow, more atrophied face” on the same side as the tilt (opposite the palsy). They quantified the asymmetry using patient photographs with lines drawn across the eyes and mouth (Fig 1). Paysee et al3 showed similar results, describing such changes as “facial hemihypoplasia.” Their patients included one with a paradoxical heat tilt to the same side as the superior oblique palsy. The
Originally received: November 8, 1998. Accepted: August 30, 1999. Manuscript no. 98541. Section of Pediatric Ophthalmology, Eye Consultants of Atlanta, Scottish Rite Children’s Medical Center, Atlanta, Georgia. Presented in part at the 1998 Annual Meeting of the American Academy of Ophthalmology, New Orleans, Louisiana, November, 1998. Supported in part by a grant from the James Hall Foundation. The authors acknowledge no financial interest in the subjects of the paper. Reprint requests to Marc F. Greenberg, MD, 5455 Meridian Mark Road, Suite 220, Atlanta, GA 30342. © 2000 by the American Academy of Ophthalmology Published by Elsevier Science Inc.
facial hypoplasia was still on the side of the tilt, suggesting that the tilt, rather than the palsy, may be inducing the facial asymmetry. Goodman et al4 have suggested that “deformational plagiocephaly” is the cause of the facial asymmetry. “Plagiocephaly” is a general term for skull and facial asymmetry derived from the Greek terms plagios, which means “oblique” or “slanted,” and kephale or “head.” In deformational plagiocephaly, perhaps the most common form of skull and facial asymmetry, the occiput and the frontal bone and full face become deformed by intrauterine or postnatal positioning on one side of the infants’s head.5 The authors postulated that infantile skull deformation from preferential sleeping on one side was the likely mechanism of asymmetry. Nevertheless, the exact cause of the facial asymmetry in ocular torticollis has not been established. Certain strabismus syndromes, such as craniosynostosis (premature fusion of cranial sutures),6,7 or Goldenhar’s syndrome (hemifacial microsomia) featuring Duane’s syndrome8 may have skull or facial asymmetry as a primary feature. However, most cases of ocular torticollis are not associated with any known plagiocephaly syndromes. To study the relationship between strabismus and facial asymmetry, we began to observe patients with many forms of ocular torticollis (Table 1). ISSN 0161-6420/00/$–see front matter PII S0161-6420(99)00004-4
173
Ophthalmology Volume 107, Number 1, January 2000 Table 1. Causes of Horizontal Ocular Torticollis (Head Turns or Tilts) Cranial nerve III, IV, or VI palsy Isolated extraocular muscle palsy Extraocular muscle restriction Thyroid Trauma Congenital Nystagmus Dissociated strabismus complex Duane’s syndrome Brown’s syndrome Refractive errors Congenital homonymous hemianopia Monocular blindness Supranuclear (tilt reaction) Macular heterotopia
Methods
Figure 1. Right superior oblique palsy with left head turn and tilt. Lines drawn between pupils and mouth angles often converge toward the side of torticollis in patient with facial asymmetry.
Patients with ocular torticollis in one author’s (MFG) private practice were observed for facial asymmetry from July 1995 to August 1998. Patients with pure vertical torticollis, as caused by ptosis or A and V patterns were excluded because chin-up or chin-down positions would not be expected to create left-right facial asymmetries. Fifty-three selected patients were photographed to best illustrate the asymmetries observed. From January
Figure 2. A, Congenital left superior oblique palsy. Patient usually assumes right head tilt and turn. With torticollis eliminated, frontal, maxillary, and mandibular facial areas appear compressed from the right side or reduced in mass on patient’s right side. B, With chin up, right facial compression or reduced mass is well seen. Also, note left deviation of nasal tip (upper arrow) and columella (lower arrow).
174
Greenberg and Pollard 䡠 Ocular Plagiocephaly 1998 to August 1998 all 44 patients with horizontal ocular torticollis seen by the same author were recorded. Specific information regarding the direction of head turn or tilt and the side of any appearance of facial compression or reduced facial mass as judged by the author’s observation was noted. One patient with acute superior oblique palsy was excluded because no facial asymmetry was believed to be present. Patients’ faces were viewed (1) with the patients looking straight ahead in their assumed head position, (2) with the head positioned straight ahead, (3) with the chin down, and (4) with the chin up. Straight ahead was judged when the patients eyes were centered in the palpebral fissures. Three areas of the face evaluated included (1) the frontal, above eye level, (2) the maxillary, below eye level and above mouth level, and (3) the mandibular, below mouth level. A review of all 53 patients photographed was made by the same author to document the direction of any horizontal deviation of the nasal tip and columella (Fig 2).
Results The numbers and types of ocular torticollis recorded from January to July 1998 are presented in Table 2. Ten different forms of ocular torticollis were documented. Forty-one of 43 patients had an appearance of facial compression or reduced volume in all facial areas on the same side as the face turn or tilt (Figs 2 to 6). Patients with large-angle, constant torticollis, such as in superior oblique palsy or Duane’s syndrome, had the more prominent asymmetries; those with intermittent torticollis, such as in nystagmus or dissociated deviations, had more subtle findings. One patient with dissociated vertical deviation and another with Duane’s syndrome had appearance of facial compression opposite to a face turn. The patient with Duane’s syndrome did have a father who had similar facial asymmetry but no strabismus. One patient with inferior oblique palsy had the turn and tilt in opposite directions (Fig 4). This patient and two others with similar palsies photographed before the consecutive clinical series had the reduced facial volume on the side of the tilt, and this was considered the direction of the torticollis. The results of the nasal tip and columella deviation with respect to the direction of face turn are shown in Table 3. Six of the eight nonsuperior oblique palsy patients with nasal deviation toward the side of the torticollis had head tilts. One patient with Duane’s syndrome and one with nystagmus had only face turns. All 17 nonsuperior oblique palsy patients with nasal deviation opposite the side of torticollis had pure face turns. Superior oblique palsy patients were almost evenly split, with eight having nasal deviation to the side of the torticollis and nine having deviation opposite. Statistical analysis was not performed because of small sample size.
Figure 3. Congenital nystagmus with left face turn (not shown). Chin-up view shows appearance of compression from the patient’s left or reduced mass of all left facial areas. Note mandibular asymmetry.
Conclusion In this article, we report facial asymmetry, with an appearance of facial compression or reduced facial mass, ipsilateral to the head turn or tilt in patients with a wide variety of types of ocular torticollis, including face turns. Because similar findings are present in many types of ocular torticollis, a unifying cause seems likely, and we postulate that the facial asymmetry in all types may be caused by the tilt or turn itself. Clearly, the unmasked, subjective observations of one physician do not prove to the reader that such facial asymmetries exist. Nevertheless, the reader is not asked to accept these observations without objective data
Table 2. Side of Torticollis vs. Side of Facial Compression or Reduced Mass
Congenital superior oblique palsy Acquired VI nerve palsy Congenital inferior oblique palsy Acquired inferior rectus restriction Nystagmus Dissociated vertical deviation Dissociated horizontal deviation Duane’s syndrome Brown’s syndrome Monocular blindness Total
Same
Opposite
14 2 1 1 7 1 4 9 1 1 41
0 0 0 0 0 1 0 1 0 0 2
Figure 4. Right inferior oblique palsy with right head tilt and left head turn. Chin-up view with head turn eliminated. Appearance of facial compression of all areas from the patient’s right or reduced right facial mass. Note right deviation of nasal tip and posterior position of right upper teeth.
175
Ophthalmology Volume 107, Number 1, January 2000
Figure 5. Right congenital Brown’s syndrome with mild left turn (not shown). Chin-up view best shows mild appearance of facial compression from the left or reduced left facial mass.
but rather is invited to observe his or her own patients for similar findings. The purpose of this article is to describe the skull and facial asymmetry we observed with ocular torticollis and to
elaborate on the concept of “ocular plagiocephaly,” in which eye muscle imbalance could cause skull and facial asymmetry. Other forms of plagiocephaly, such as deformational or craniosynostosis, start with abnormal bone and muscle configurations at or shortly after birth and are frequently accompanied by torticollis. Such primary musculoskeletal abnormalities could cause the plagiocephaly, the torticollis, or both. In “ocular plagiocephaly,” however, the bones and muscles are normal at birth; the head tilt or turn is purely for neurosensory reasons and so the effect of prolonged torticollis on facial growth can be purely judged. We would like to suggest that the facial asymmetry in ocular torticollis may be caused by differential remodeling of the head and face as a result of the turn or tilt. Bone, muscle, fat, and skin are all living tissues and undergo continuous turnover and remodeling throughout life. It has been estimated that the human adult skeleton turns over almost 20% of its bone mass per year9 or complete turnover in 5 years. Presumably, normal growth in children and increased tissue resorption and atrophy in the elderly could lead to even more rapid remodeling changes in these age groups. Other known examples of long-term tissue restructuring would include remodeling of extraocular muscle sarcomeres as a mechanism for contracture of these muscles in strabismus10 and increasing arm length of tennis players on the side of the stroke arm.11 Evidence exists that pattern development of the face is under genetic control and that the same genes active in the young during development continue to express themselves during adulthood.12 Such control must be influenced by a head tilt or turn. The mechanisms by which a head turn or tilt could affect facial remodeling are not known. Constant muscle tension or changes in muscle innervation could play a role; however, the muscle responsible for much of the head turn or tilt, the sternocleidomastoid, inserts behind the ear, distant from the face and forehead and with no connection to the mandible. Furthermore, in face turns, this muscle acts on the side opposite the turn and facial compression. Gravity may play a role because inferior oblique palsy patients with turns opposite their tilts showed atrophy toward the side of the tilt. Furthermore, Wilson and Hoxie2 noted the tip of the nose turning toward the side of a head tilt in superior oblique palsy, and gravity would seem a Table 3. Photo Review—53 Patients with Ocular Torticollis (Direction of Nasal Tip and Columella Deviation vs. Direction of Head Tilt or Turn)
Figure 6. A, Left VI nerve palsy acquired 20 years before photo with large left face turn (not shown). Face front view with appearance of facial compression from the patient’s left or reduced left facial mass. Note atrophy of left temple and increased depth of skin wrinkles around the eye, the nasolabial fold, the cheek dimple, and the chin on the left side. B, Chin-down view shows extreme right nasal tip deviation.
176
Duane’s syndrome Brown’s syndrome Nystagmus Acquired VI nerve palsy Dissociative horizontal deviation Superior oblique palsy Inferior oblique palsy Dissociative vertical deviation Acquired inferior rectus restriction with ipsilateral tilt Total
Same
Opposite
No Deviation
2 0 1 0 0 9 3 1
10 2 2 1 2 8 0 0
2 2 4 0 0 2 0 1
1 17
0 25
0 11
Greenberg and Pollard 䡠 Ocular Plagiocephaly likely factor. When such patients are viewed in their preferred tilted position, the nose tip often points straight down toward the ground. In our photo review, patients with nasal deviation toward the side of their torticollis predominantly had head tilts. In contrast, the nose tip and columella pointed away from the turn in most patients with pure face turns, such as in Duane’s syndrome. This included one patient with adultonset acquired VI nerve palsy. In patients with such face turns, genetic pattern influences must operate to keep the nose pointing straight forward with respect to the body and not the turned head. In this view, the appearance of compression in the frontal bone, maxilla, and mandible in face turn ocular torticollis can also be thought of as changes intended to keep the face oriented forward with respect to the body, despite the turned head. We are unaware of previous reports describing facial asymmetry in face turn ocular torticollis. Interestingly, our observation that the nasal tip tends to point toward a head tilt but away from a turn may explain our mixed results in superior oblique nasal tip deviation (Table 3). In superior oblique palsy, the head tilt and turn are commonly in the same direction, producing a conflict in predicting nasal tip deviation—toward the tilt or away from the turn. In cases of superior oblique palsy in which the nasal tip points away from the torticollis, we assume the face turn is more significant than the tilt. The asymmetries we describe are three dimensional, often requiring different views to appreciate, and we have yet to settle on an acceptable objective office measuring method. Wilson and Hoxie2 and others3,4 have used unstandardized photographs with pupil-to-pupil and mouth angle–to–mouth angle reference lines to illustrate asymmetry (Fig 1). We believe this method is insufficient to detect more subtle asymmetries that we observed clinically, such as in dissociated deviations. In addition, we have observed many superior oblique palsy patients with marked facial asymmetry when viewed from all angles, whose asymmetry would be unappreciated by such methods (such as in Fig 2). Other photographic measuring methods have been described13; however, after years of photographing patients with facial asymmetry, it is our feeling that such twodimensional representations of the three-dimensional face are fraught with distortions. The photographs in this article were chosen because they best represent the clinical appearance we observed; however, our experience suggests they are not adequate for measurement purposes. The author’s determination of the laterality of facial asymmetry was done during clinical examinations rather than from patient photographs for this reason. We assessed direction of nasal tip and columella deviation photographically only because we became interested in this particular facet of facial asymmetry after our consecutive series of examinations was completed. Tulasne and Tessier,14 two of the pioneers in classifying plagiocephaly, have written, “Judging the correct orientation of plagiocephaly cannot be done on a dry skull or on films . . . correct examination . . . can only be undertaken in front of the patient,” and we agree. Familiarity with the appearance of “ocular plagioceph-
aly” is useful to the clinical ophthalmologist as a sign of torticollis chronicity. In acute head or facial trauma cases with strabismus, correctly oriented facial asymmetry may suggest a pre-existing motility defect. Likewise, in cases of extraocular muscle palsy, characteristic facial asymmetry can occasionally help obviate the need for extensive workup. In cases of null point nystagmus, such asymmetry can support the indication for Kestenbaum procedures. In addition, facial asymmetry may be the only clinical sign that can be compared with office notes to confirm the correct direction of eye muscle surgery when the patient is anesthetized just before an operation. Please note that facial asymmetry cannot be used in undiagnosed torticollis to suggest an ocular cause because a head tilt or turn from any other cause should still lead to the same asymmetries. Furthermore, even in cases of known ocular torticollis, it cannot be considered a pathognomonic sign because two or more other factors influencing facial asymmetry may also exist simultaneously. If, for example, deformational plagiocephaly were present on the opposite side of a patient with an ocular face turn, the facial reduced volume or appearance of compression might be opposite to the turn if the deformation was more significant than the torticollis. In other cases, familial facial asymmetries may play a role, as possibly occurred in our single patient with Duane’s syndrome with asymmetry opposite his face turn. Thus, the few patients we encountered with asymmetry contralateral to their turn or tilt may have other components of facial asymmetry more dominant than the ocular plagiocephaly factors. As mentioned, Goodman et al4 believed facial asymmetry in superior oblique palsy was a result of deformational plagiocephaly caused by abnormal sleep positioning. They suggested that early strabismus surgery, before skull suture closure, might prevent this complication. Paysee et al3 have noted that ocular torticollis disappears in the supine position and during sleep and would therefore not cause deformational plagiocephaly. Our observations of facial asymmetry in toddler-onset dissociated vertical deviation and in longstanding adult-onset ocular torticollis would also suggest that infantile skull deformation is not the cause of the asymmetry. Although we agree that strabismus surgery to prevent long-term remodeling asymmetries might best achieve this goal the earlier it is undertaken, we are uncertain, however, of any critical time limit in which surgery must take place. The finding of facial asymmetries in acquired adult-onset ocular torticollis would suggest some hope that strabismus surgery at any age could improve, or at least lessen the further progression of, facial asymmetries if it alters the torticollis. It is commonly observed that the turn or tilt in ocular torticollis sometimes persists despite surgical resolution of the strabismus. Residual torticollis could presumably continue to influence the same long-term asymmetric facial remodeling changes as it did before strabismus surgery. Lastly, the ability to alter facial asymmetry by reducing ocular torticollis is presumed, and such benefits of strabismus surgery have yet to be shown conclusively. Also, many of the patients we observed with plagiocephaly and ocular torticollis were unhindered by or even unaware of their
177
Ophthalmology Volume 107, Number 1, January 2000 facial asymmetry. Strabismus surgeons will have to decide individually whether to list facial asymmetry from ocular torticollis as one of their indications to operate. In conclusion, we present our observations that patients with ocular torticollis from a variety of causes, including face turns, share a similar appearance of facial compression or reduced facial mass on the side of the turn or tilt. Patients with head tilts often show nasal deviation toward the tilt; those with head turns often show nasal deviation away from the turn. We believe the torticollis may cause the facial asymmetry by altering the continuous tissue remodeling process and refer to such patients as having “ocular plagiocephaly.” In the presence of strabismus, such facial asymmetry is a useful clinical sign of chronicity of the eye muscle imbalance. Strabismus surgery that alters the torticollis could alter the asymmetric remodeling process for the better.
References 1. Parks MM. Isolated cyclovertical muscle palsy. Arch Ophthalmol 1958;60:1027–35. 2. Wilson ME, Hoxie J. Facial asymmetry in superior oblique muscle palsy. J Pediatr Ophthalmol Strabismus 1993;30:315– 18. 3. Paysee EA, Coats DK, Plager DA. Facial asymmetry and tendon laxity in superior oblique palsy. J Pediatr Ophthalmol Strabismus 1995;32:158 – 61. 4. Goodman CR, Chabner E, Guyton DL. Should early strabismus surgery be performed for ocular torticollis to prevent facial asymmetry? J Pediatr Ophthalmol Strabismus 1995;32: 162– 6.
5. Bruneteau RJ, Mulliken JB. Frontal plagiocephaly: synostotic, compensational, or deformational. Plast Reconstr Surg 1992; 89:21–31; discussion 32–3. 6. Bagolini B, Campos EC, Chiesi C. Plagiocephaly causing superior oblique deficiency and ocular torticollis: a new clinical entity. Arch Ophthalmol 1982;100:1093– 6. 7. Robb RM, Boger WP III. Vertical strabismus associated with plagiocephaly. J Pediatr Ophthalmol Strabismus 1983;20:58 – 62. 8. Miller MT. Association of Duane retraction syndrome with craniofacial malformations. J Craniofac Genet Dev Biol Suppl 1985;1:273– 82. 9. Holick MF, Krane SM, Potts JT Jr. Calcium, phosphorus and bone metabolism: calcium regulating hormones. In: Isselbacher KJ, Braunwald E, Wilson JD, eds. Harrison’s Principles of Internal Medicine, 13th ed. Vol. 2. New York: McGraw-Hill, 1994;1857. 10. Scott AB. Change of eye muscle sarcomeres according to eye postion. J Pediatr Ophthalmol Strabismus 1994;31:85– 8. 11. Krahl H, Michaelis U, Pieper HG, et al. Stimulation of bone growth through sports: a radiologic investigation of the upper extremeties in professional tennis players. Am J Sports Med 1994;22:751–7. 12. Zernik JH, Nowroozi N, Liu YH, Maxson R. Development, maturation, and aging of the alveolar bone: new insights. Dent Clin North Am 1997;41(1):1–15. 13. Ferguson JW. Cephalometric interpretation and assessment of facial asymmetry secondary to congenital torticollis: the significance of cranial base reference lines. Int J Oral Maxillofac Surg 1993;22:7–10. 14. Tulasne JF, Tessier P. Analysis and late treatment of plagiocephaly. Unilateral coronal synostosis. Scand J Plast Reconstr Surg 1981;15:257– 63.
Discussion by Bruce M. Schnall, MD Facial asymmetry or plagiocephaly related to head tilt with a chronic superior oblique palsy has been reported by a number of authors.1– 4 It is typically described as the midportion of the face being smaller on the side to which the head is tilted.1– 4 It is usually associated with the root of the nose being displaced in the direction of the head tilt. Associated mandibular asymmetry can cause dental malocclusion.4,5 Similar facial asymmetry has been reported with head tilts caused by other congenital abnormalities such as congenital muscular torticollis resulting from a structural deformity of the sternocleidomastoid muscle.6 Dr. Greenberg looked at two aspects of facial asymmetry in two overlapping groups of patients by two different methods. The first is a prospective study of 44 consecutive patients who were seen by the author with head tilt or head turn over an 8-month period. The method was subjective examination of the face, viewing it with the chin-down, chin-up, and straight-ahead position. Of these 44 patients, 14 had superior oblique palsy. We can compare the results reported in this article to those of similar series.
From the Department of Pediatric Ophthalmology, Wills Eye Hospital, Philadelphia, Pennsylvania. Address correspondence to Bruce M. Schnall, MD, Department of Pediatric Ophthalmology, Wills Eye Hospital, 900 Walnut Street, Philadelphia, PA 19107.
178
Fourteen of 14 (100%) patients with superior oblique palsy in Dr. Greenberg’s series were reported as having reduced mass of the midportion of the face on the side of the head tilt. Wilson and Hoxie2 reported that seven of nine (77%) patients with congenital superior oblique palsy in their series had facial asymmetry, whereas Pasyee et al3 reported that 16 of 21 (76%) patients in their series with superior oblique palsy had facial asymmetry. Why did Dr. Greenberg find a higher percentage of his patients with facial asymmetry? It may be related to his methods. Drs. Wilson and Paysee relied on photographs in the frontal view to determine whether plagiocephaly was present. Dr. Greenberg made his assessment while examining the patient and by viewing the patient in several different planes. I suspect that this method is more sensitive, allowing Dr. Greenberg to detect smaller degrees of facial asymmetry. Dr. Greenberg also detected facial asymmetry in patients with chronic head turn associated with certain eye muscle disorders. To my knowledge, with the exception of the association of hemifacial microsomia with Duane’s syndrome,7 this has not been reported. Other investigators will need to validate this observation. The second part of this study involved a review of photographs of 53 selected patients with torticollis to determine whether the nasal tip and columella (nasal root) were displaced. They observed a tendency for the nose to be deviated toward the head tilt, which is consistent with what other authors have reported.2 In patients
Originally received: November 8, 1998. Accepted: August 30, 1999. Manuscript no. 98541. Section of Pediatric Ophthalmology, Eye Consultants of Atlanta, Scottish Rite Children’s Medical Center, Atlanta, Georgia. Presented in part at the 1998 Annual Meeting of the American Academy of Ophthalmology, New Orleans, Louisiana, November, 1998. Supported in part by a grant from the James Hall Foundation. The authors acknowledge no financial interest in the subjects of the paper. Reprint requests to Marc F. Greenberg, MD, 5455 Meridian Mark Road, Suite 220, Atlanta, GA 30342. © 2000 by the American Academy of Ophthalmology Published by Elsevier Science Inc.
facial hypoplasia was still on the side of the tilt, suggesting that the tilt, rather than the palsy, may be inducing the facial asymmetry. Goodman et al4 have suggested that “deformational plagiocephaly” is the cause of the facial asymmetry. “Plagiocephaly” is a general term for skull and facial asymmetry derived from the Greek terms plagios, which means “oblique” or “slanted,” and kephale or “head.” In deformational plagiocephaly, perhaps the most common form of skull and facial asymmetry, the occiput and the frontal bone and full face become deformed by intrauterine or postnatal positioning on one side of the infants’s head.5 The authors postulated that infantile skull deformation from preferential sleeping on one side was the likely mechanism of asymmetry. Nevertheless, the exact cause of the facial asymmetry in ocular torticollis has not been established. Certain strabismus syndromes, such as craniosynostosis (premature fusion of cranial sutures),6,7 or Goldenhar’s syndrome (hemifacial microsomia) featuring Duane’s syndrome8 may have skull or facial asymmetry as a primary feature. However, most cases of ocular torticollis are not associated with any known plagiocephaly syndromes. To study the relationship between strabismus and facial asymmetry, we began to observe patients with many forms of ocular torticollis (Table 1). ISSN 0161-6420/00/$–see front matter PII S0161-6420(99)00004-4
173
Ophthalmology Volume 107, Number 1, January 2000 Table 1. Causes of Horizontal Ocular Torticollis (Head Turns or Tilts) Cranial nerve III, IV, or VI palsy Isolated extraocular muscle palsy Extraocular muscle restriction Thyroid Trauma Congenital Nystagmus Dissociated strabismus complex Duane’s syndrome Brown’s syndrome Refractive errors Congenital homonymous hemianopia Monocular blindness Supranuclear (tilt reaction) Macular heterotopia
Methods
Figure 1. Right superior oblique palsy with left head turn and tilt. Lines drawn between pupils and mouth angles often converge toward the side of torticollis in patient with facial asymmetry.
Patients with ocular torticollis in one author’s (MFG) private practice were observed for facial asymmetry from July 1995 to August 1998. Patients with pure vertical torticollis, as caused by ptosis or A and V patterns were excluded because chin-up or chin-down positions would not be expected to create left-right facial asymmetries. Fifty-three selected patients were photographed to best illustrate the asymmetries observed. From January
Figure 2. A, Congenital left superior oblique palsy. Patient usually assumes right head tilt and turn. With torticollis eliminated, frontal, maxillary, and mandibular facial areas appear compressed from the right side or reduced in mass on patient’s right side. B, With chin up, right facial compression or reduced mass is well seen. Also, note left deviation of nasal tip (upper arrow) and columella (lower arrow).
174
Greenberg and Pollard 䡠 Ocular Plagiocephaly 1998 to August 1998 all 44 patients with horizontal ocular torticollis seen by the same author were recorded. Specific information regarding the direction of head turn or tilt and the side of any appearance of facial compression or reduced facial mass as judged by the author’s observation was noted. One patient with acute superior oblique palsy was excluded because no facial asymmetry was believed to be present. Patients’ faces were viewed (1) with the patients looking straight ahead in their assumed head position, (2) with the head positioned straight ahead, (3) with the chin down, and (4) with the chin up. Straight ahead was judged when the patients eyes were centered in the palpebral fissures. Three areas of the face evaluated included (1) the frontal, above eye level, (2) the maxillary, below eye level and above mouth level, and (3) the mandibular, below mouth level. A review of all 53 patients photographed was made by the same author to document the direction of any horizontal deviation of the nasal tip and columella (Fig 2).
Results The numbers and types of ocular torticollis recorded from January to July 1998 are presented in Table 2. Ten different forms of ocular torticollis were documented. Forty-one of 43 patients had an appearance of facial compression or reduced volume in all facial areas on the same side as the face turn or tilt (Figs 2 to 6). Patients with large-angle, constant torticollis, such as in superior oblique palsy or Duane’s syndrome, had the more prominent asymmetries; those with intermittent torticollis, such as in nystagmus or dissociated deviations, had more subtle findings. One patient with dissociated vertical deviation and another with Duane’s syndrome had appearance of facial compression opposite to a face turn. The patient with Duane’s syndrome did have a father who had similar facial asymmetry but no strabismus. One patient with inferior oblique palsy had the turn and tilt in opposite directions (Fig 4). This patient and two others with similar palsies photographed before the consecutive clinical series had the reduced facial volume on the side of the tilt, and this was considered the direction of the torticollis. The results of the nasal tip and columella deviation with respect to the direction of face turn are shown in Table 3. Six of the eight nonsuperior oblique palsy patients with nasal deviation toward the side of the torticollis had head tilts. One patient with Duane’s syndrome and one with nystagmus had only face turns. All 17 nonsuperior oblique palsy patients with nasal deviation opposite the side of torticollis had pure face turns. Superior oblique palsy patients were almost evenly split, with eight having nasal deviation to the side of the torticollis and nine having deviation opposite. Statistical analysis was not performed because of small sample size.
Figure 3. Congenital nystagmus with left face turn (not shown). Chin-up view shows appearance of compression from the patient’s left or reduced mass of all left facial areas. Note mandibular asymmetry.
Conclusion In this article, we report facial asymmetry, with an appearance of facial compression or reduced facial mass, ipsilateral to the head turn or tilt in patients with a wide variety of types of ocular torticollis, including face turns. Because similar findings are present in many types of ocular torticollis, a unifying cause seems likely, and we postulate that the facial asymmetry in all types may be caused by the tilt or turn itself. Clearly, the unmasked, subjective observations of one physician do not prove to the reader that such facial asymmetries exist. Nevertheless, the reader is not asked to accept these observations without objective data
Table 2. Side of Torticollis vs. Side of Facial Compression or Reduced Mass
Congenital superior oblique palsy Acquired VI nerve palsy Congenital inferior oblique palsy Acquired inferior rectus restriction Nystagmus Dissociated vertical deviation Dissociated horizontal deviation Duane’s syndrome Brown’s syndrome Monocular blindness Total
Same
Opposite
14 2 1 1 7 1 4 9 1 1 41
0 0 0 0 0 1 0 1 0 0 2
Figure 4. Right inferior oblique palsy with right head tilt and left head turn. Chin-up view with head turn eliminated. Appearance of facial compression of all areas from the patient’s right or reduced right facial mass. Note right deviation of nasal tip and posterior position of right upper teeth.
175
Ophthalmology Volume 107, Number 1, January 2000
Figure 5. Right congenital Brown’s syndrome with mild left turn (not shown). Chin-up view best shows mild appearance of facial compression from the left or reduced left facial mass.
but rather is invited to observe his or her own patients for similar findings. The purpose of this article is to describe the skull and facial asymmetry we observed with ocular torticollis and to
elaborate on the concept of “ocular plagiocephaly,” in which eye muscle imbalance could cause skull and facial asymmetry. Other forms of plagiocephaly, such as deformational or craniosynostosis, start with abnormal bone and muscle configurations at or shortly after birth and are frequently accompanied by torticollis. Such primary musculoskeletal abnormalities could cause the plagiocephaly, the torticollis, or both. In “ocular plagiocephaly,” however, the bones and muscles are normal at birth; the head tilt or turn is purely for neurosensory reasons and so the effect of prolonged torticollis on facial growth can be purely judged. We would like to suggest that the facial asymmetry in ocular torticollis may be caused by differential remodeling of the head and face as a result of the turn or tilt. Bone, muscle, fat, and skin are all living tissues and undergo continuous turnover and remodeling throughout life. It has been estimated that the human adult skeleton turns over almost 20% of its bone mass per year9 or complete turnover in 5 years. Presumably, normal growth in children and increased tissue resorption and atrophy in the elderly could lead to even more rapid remodeling changes in these age groups. Other known examples of long-term tissue restructuring would include remodeling of extraocular muscle sarcomeres as a mechanism for contracture of these muscles in strabismus10 and increasing arm length of tennis players on the side of the stroke arm.11 Evidence exists that pattern development of the face is under genetic control and that the same genes active in the young during development continue to express themselves during adulthood.12 Such control must be influenced by a head tilt or turn. The mechanisms by which a head turn or tilt could affect facial remodeling are not known. Constant muscle tension or changes in muscle innervation could play a role; however, the muscle responsible for much of the head turn or tilt, the sternocleidomastoid, inserts behind the ear, distant from the face and forehead and with no connection to the mandible. Furthermore, in face turns, this muscle acts on the side opposite the turn and facial compression. Gravity may play a role because inferior oblique palsy patients with turns opposite their tilts showed atrophy toward the side of the tilt. Furthermore, Wilson and Hoxie2 noted the tip of the nose turning toward the side of a head tilt in superior oblique palsy, and gravity would seem a Table 3. Photo Review—53 Patients with Ocular Torticollis (Direction of Nasal Tip and Columella Deviation vs. Direction of Head Tilt or Turn)
Figure 6. A, Left VI nerve palsy acquired 20 years before photo with large left face turn (not shown). Face front view with appearance of facial compression from the patient’s left or reduced left facial mass. Note atrophy of left temple and increased depth of skin wrinkles around the eye, the nasolabial fold, the cheek dimple, and the chin on the left side. B, Chin-down view shows extreme right nasal tip deviation.
176
Duane’s syndrome Brown’s syndrome Nystagmus Acquired VI nerve palsy Dissociative horizontal deviation Superior oblique palsy Inferior oblique palsy Dissociative vertical deviation Acquired inferior rectus restriction with ipsilateral tilt Total
Same
Opposite
No Deviation
2 0 1 0 0 9 3 1
10 2 2 1 2 8 0 0
2 2 4 0 0 2 0 1
1 17
0 25
0 11
Greenberg and Pollard 䡠 Ocular Plagiocephaly likely factor. When such patients are viewed in their preferred tilted position, the nose tip often points straight down toward the ground. In our photo review, patients with nasal deviation toward the side of their torticollis predominantly had head tilts. In contrast, the nose tip and columella pointed away from the turn in most patients with pure face turns, such as in Duane’s syndrome. This included one patient with adultonset acquired VI nerve palsy. In patients with such face turns, genetic pattern influences must operate to keep the nose pointing straight forward with respect to the body and not the turned head. In this view, the appearance of compression in the frontal bone, maxilla, and mandible in face turn ocular torticollis can also be thought of as changes intended to keep the face oriented forward with respect to the body, despite the turned head. We are unaware of previous reports describing facial asymmetry in face turn ocular torticollis. Interestingly, our observation that the nasal tip tends to point toward a head tilt but away from a turn may explain our mixed results in superior oblique nasal tip deviation (Table 3). In superior oblique palsy, the head tilt and turn are commonly in the same direction, producing a conflict in predicting nasal tip deviation—toward the tilt or away from the turn. In cases of superior oblique palsy in which the nasal tip points away from the torticollis, we assume the face turn is more significant than the tilt. The asymmetries we describe are three dimensional, often requiring different views to appreciate, and we have yet to settle on an acceptable objective office measuring method. Wilson and Hoxie2 and others3,4 have used unstandardized photographs with pupil-to-pupil and mouth angle–to–mouth angle reference lines to illustrate asymmetry (Fig 1). We believe this method is insufficient to detect more subtle asymmetries that we observed clinically, such as in dissociated deviations. In addition, we have observed many superior oblique palsy patients with marked facial asymmetry when viewed from all angles, whose asymmetry would be unappreciated by such methods (such as in Fig 2). Other photographic measuring methods have been described13; however, after years of photographing patients with facial asymmetry, it is our feeling that such twodimensional representations of the three-dimensional face are fraught with distortions. The photographs in this article were chosen because they best represent the clinical appearance we observed; however, our experience suggests they are not adequate for measurement purposes. The author’s determination of the laterality of facial asymmetry was done during clinical examinations rather than from patient photographs for this reason. We assessed direction of nasal tip and columella deviation photographically only because we became interested in this particular facet of facial asymmetry after our consecutive series of examinations was completed. Tulasne and Tessier,14 two of the pioneers in classifying plagiocephaly, have written, “Judging the correct orientation of plagiocephaly cannot be done on a dry skull or on films . . . correct examination . . . can only be undertaken in front of the patient,” and we agree. Familiarity with the appearance of “ocular plagioceph-
aly” is useful to the clinical ophthalmologist as a sign of torticollis chronicity. In acute head or facial trauma cases with strabismus, correctly oriented facial asymmetry may suggest a pre-existing motility defect. Likewise, in cases of extraocular muscle palsy, characteristic facial asymmetry can occasionally help obviate the need for extensive workup. In cases of null point nystagmus, such asymmetry can support the indication for Kestenbaum procedures. In addition, facial asymmetry may be the only clinical sign that can be compared with office notes to confirm the correct direction of eye muscle surgery when the patient is anesthetized just before an operation. Please note that facial asymmetry cannot be used in undiagnosed torticollis to suggest an ocular cause because a head tilt or turn from any other cause should still lead to the same asymmetries. Furthermore, even in cases of known ocular torticollis, it cannot be considered a pathognomonic sign because two or more other factors influencing facial asymmetry may also exist simultaneously. If, for example, deformational plagiocephaly were present on the opposite side of a patient with an ocular face turn, the facial reduced volume or appearance of compression might be opposite to the turn if the deformation was more significant than the torticollis. In other cases, familial facial asymmetries may play a role, as possibly occurred in our single patient with Duane’s syndrome with asymmetry opposite his face turn. Thus, the few patients we encountered with asymmetry contralateral to their turn or tilt may have other components of facial asymmetry more dominant than the ocular plagiocephaly factors. As mentioned, Goodman et al4 believed facial asymmetry in superior oblique palsy was a result of deformational plagiocephaly caused by abnormal sleep positioning. They suggested that early strabismus surgery, before skull suture closure, might prevent this complication. Paysee et al3 have noted that ocular torticollis disappears in the supine position and during sleep and would therefore not cause deformational plagiocephaly. Our observations of facial asymmetry in toddler-onset dissociated vertical deviation and in longstanding adult-onset ocular torticollis would also suggest that infantile skull deformation is not the cause of the asymmetry. Although we agree that strabismus surgery to prevent long-term remodeling asymmetries might best achieve this goal the earlier it is undertaken, we are uncertain, however, of any critical time limit in which surgery must take place. The finding of facial asymmetries in acquired adult-onset ocular torticollis would suggest some hope that strabismus surgery at any age could improve, or at least lessen the further progression of, facial asymmetries if it alters the torticollis. It is commonly observed that the turn or tilt in ocular torticollis sometimes persists despite surgical resolution of the strabismus. Residual torticollis could presumably continue to influence the same long-term asymmetric facial remodeling changes as it did before strabismus surgery. Lastly, the ability to alter facial asymmetry by reducing ocular torticollis is presumed, and such benefits of strabismus surgery have yet to be shown conclusively. Also, many of the patients we observed with plagiocephaly and ocular torticollis were unhindered by or even unaware of their
177
Ophthalmology Volume 107, Number 1, January 2000 facial asymmetry. Strabismus surgeons will have to decide individually whether to list facial asymmetry from ocular torticollis as one of their indications to operate. In conclusion, we present our observations that patients with ocular torticollis from a variety of causes, including face turns, share a similar appearance of facial compression or reduced facial mass on the side of the turn or tilt. Patients with head tilts often show nasal deviation toward the tilt; those with head turns often show nasal deviation away from the turn. We believe the torticollis may cause the facial asymmetry by altering the continuous tissue remodeling process and refer to such patients as having “ocular plagiocephaly.” In the presence of strabismus, such facial asymmetry is a useful clinical sign of chronicity of the eye muscle imbalance. Strabismus surgery that alters the torticollis could alter the asymmetric remodeling process for the better.
References 1. Parks MM. Isolated cyclovertical muscle palsy. Arch Ophthalmol 1958;60:1027–35. 2. Wilson ME, Hoxie J. Facial asymmetry in superior oblique muscle palsy. J Pediatr Ophthalmol Strabismus 1993;30:315– 18. 3. Paysee EA, Coats DK, Plager DA. Facial asymmetry and tendon laxity in superior oblique palsy. J Pediatr Ophthalmol Strabismus 1995;32:158 – 61. 4. Goodman CR, Chabner E, Guyton DL. Should early strabismus surgery be performed for ocular torticollis to prevent facial asymmetry? J Pediatr Ophthalmol Strabismus 1995;32: 162– 6.
5. Bruneteau RJ, Mulliken JB. Frontal plagiocephaly: synostotic, compensational, or deformational. Plast Reconstr Surg 1992; 89:21–31; discussion 32–3. 6. Bagolini B, Campos EC, Chiesi C. Plagiocephaly causing superior oblique deficiency and ocular torticollis: a new clinical entity. Arch Ophthalmol 1982;100:1093– 6. 7. Robb RM, Boger WP III. Vertical strabismus associated with plagiocephaly. J Pediatr Ophthalmol Strabismus 1983;20:58 – 62. 8. Miller MT. Association of Duane retraction syndrome with craniofacial malformations. J Craniofac Genet Dev Biol Suppl 1985;1:273– 82. 9. Holick MF, Krane SM, Potts JT Jr. Calcium, phosphorus and bone metabolism: calcium regulating hormones. In: Isselbacher KJ, Braunwald E, Wilson JD, eds. Harrison’s Principles of Internal Medicine, 13th ed. Vol. 2. New York: McGraw-Hill, 1994;1857. 10. Scott AB. Change of eye muscle sarcomeres according to eye postion. J Pediatr Ophthalmol Strabismus 1994;31:85– 8. 11. Krahl H, Michaelis U, Pieper HG, et al. Stimulation of bone growth through sports: a radiologic investigation of the upper extremeties in professional tennis players. Am J Sports Med 1994;22:751–7. 12. Zernik JH, Nowroozi N, Liu YH, Maxson R. Development, maturation, and aging of the alveolar bone: new insights. Dent Clin North Am 1997;41(1):1–15. 13. Ferguson JW. Cephalometric interpretation and assessment of facial asymmetry secondary to congenital torticollis: the significance of cranial base reference lines. Int J Oral Maxillofac Surg 1993;22:7–10. 14. Tulasne JF, Tessier P. Analysis and late treatment of plagiocephaly. Unilateral coronal synostosis. Scand J Plast Reconstr Surg 1981;15:257– 63.
Discussion by Bruce M. Schnall, MD Facial asymmetry or plagiocephaly related to head tilt with a chronic superior oblique palsy has been reported by a number of authors.1– 4 It is typically described as the midportion of the face being smaller on the side to which the head is tilted.1– 4 It is usually associated with the root of the nose being displaced in the direction of the head tilt. Associated mandibular asymmetry can cause dental malocclusion.4,5 Similar facial asymmetry has been reported with head tilts caused by other congenital abnormalities such as congenital muscular torticollis resulting from a structural deformity of the sternocleidomastoid muscle.6 Dr. Greenberg looked at two aspects of facial asymmetry in two overlapping groups of patients by two different methods. The first is a prospective study of 44 consecutive patients who were seen by the author with head tilt or head turn over an 8-month period. The method was subjective examination of the face, viewing it with the chin-down, chin-up, and straight-ahead position. Of these 44 patients, 14 had superior oblique palsy. We can compare the results reported in this article to those of similar series.
From the Department of Pediatric Ophthalmology, Wills Eye Hospital, Philadelphia, Pennsylvania. Address correspondence to Bruce M. Schnall, MD, Department of Pediatric Ophthalmology, Wills Eye Hospital, 900 Walnut Street, Philadelphia, PA 19107.
178
Fourteen of 14 (100%) patients with superior oblique palsy in Dr. Greenberg’s series were reported as having reduced mass of the midportion of the face on the side of the head tilt. Wilson and Hoxie2 reported that seven of nine (77%) patients with congenital superior oblique palsy in their series had facial asymmetry, whereas Pasyee et al3 reported that 16 of 21 (76%) patients in their series with superior oblique palsy had facial asymmetry. Why did Dr. Greenberg find a higher percentage of his patients with facial asymmetry? It may be related to his methods. Drs. Wilson and Paysee relied on photographs in the frontal view to determine whether plagiocephaly was present. Dr. Greenberg made his assessment while examining the patient and by viewing the patient in several different planes. I suspect that this method is more sensitive, allowing Dr. Greenberg to detect smaller degrees of facial asymmetry. Dr. Greenberg also detected facial asymmetry in patients with chronic head turn associated with certain eye muscle disorders. To my knowledge, with the exception of the association of hemifacial microsomia with Duane’s syndrome,7 this has not been reported. Other investigators will need to validate this observation. The second part of this study involved a review of photographs of 53 selected patients with torticollis to determine whether the nasal tip and columella (nasal root) were displaced. They observed a tendency for the nose to be deviated toward the head tilt, which is consistent with what other authors have reported.2 In patients