Intraorbital Wood Detection by Magnetic Resonance Imaging BETH F. GREEN, MD,t STEPHEN P. KRAFT, MD,t KEITH D. CARTER, MD,2 J. RAYMOND BUNCIC, MD,t JEFFREY A. NERAD, MD/ DEREK ARMSTRONG, MD3
Abstract: The authors present two cases in which intraorbital wooden foreign bodies remained undetected after initial ophthalmologic examination and radiologic investigation which included plain orbital x-rays, orbital computed tomography (CT) scans, and, in one case, orbital ultrasound. In each case, subsequent magnetic resonance imaging (MRI) showed a well-delineated low-intensity lesion suggestive of a retained foreign body. Investigation of a case of suspected wooden foreign body in the orbit should include an MRI scan if there is no contraindication, and no foreign body has been defined on CT scan, ultrasound, or plain orbital films. Ophthalmology 1990; 97:608-611
The detection of an intraorbital wooden foreign body is difficult, particularly in cases of apparently minor trauma. Several patients have been reported in whom the usual diagnostic techniques did not detect intraorbital wood. 1-4 In each case, the trauma was described as minor, and the external injury appeared superficial. Orbital xrays rarely detect wooden fragments. 3 Computed tomography (CT) scanning can detect intraorbital wood associated with metallic paint5 or a granuloma,6 but both experimental 7 and clinical 8 studies have shown that CT has little value in detecting dry wood alone. Similarly, orbital ultrasound has had limited success in detecting intraorbital wOOd. 9 ,1O We discuss two patients in whom initial ophthalmologic examination and radiologic investigation were unable to identify intraorbital wood penetration. However, subsequent magnetic resonance imaging (MRI) in each case showed a well-delineated low-intensity intraorbital lesion suggestive of a retained foreign body. In each case a wooden foreign body was surgically removed. To our knowledge, this is the first reported use of MRI to detect intraorbital wood.
Originally received: September 28, 1989. Revision accepted: December 29, 1989. Department of Ophthalmology, The Hospital for Sick Children, Toronto. Department of Ophthalmology, University of Iowa, Iowa City. 3 Department of Radiology, The Hospital for Sick Children, Toronto. 1
2
Reprint requests to Stephen P. Kraft, MD, Department of Ophthalmology, The Hospital for Sick Children, 555 University Avenue, Toronto, Ontario, Canada M5G 1X8.
608
CASE REPORTS Case 1. A 14-year-old boy had been swinging from a tree branch which recoiled and struck his left orbit resulting in immediate total visual loss and complete ptosis. On initial examination 2 hours later, the left upper lid was swollen shut. A central stellate upper lid laceration at the level of the lid crease exposed intact orbicularis fibers with no apparent penetration site, Visual acuity was 6/6 (20/20) in the right eye and no light perception (NLP) in the left eye. Pupil sizes were 4 mm in the right eye and 7 mm in the left. The right pupil reacted briskly to light. The left pupil was unreactive to both direct light and consensual stimulation. The left eye was proptotic. Its motility was severely limited in adduction, elevation, and depression, and partially limited in abduction (Fig I). Forced ductions were positive in all planes, particularly elevation and depression. Ocular examination of the left eye revealed an inferocentral corneal abrasion and moderate flare and cells in the anterior chamber. Funduscopic examination showed moderate Berlin's edema, and the optic disc and vessels were normal. Applanation tensions were 18 mmHg in the right eye and 14 mmHg in the left. Orbital plain films (Fig 2) and CT scans (Fig 3) revealed fractures of the left orbital floor and medial wall and "air" in the superior orbit. The patient was admitted to the hospital and received intravenous cefazolin and methylprednisolone. The upper eyelid wound was explored superficially under local anesthesia, and the laceration was closed. After 3 days, there was no improvement in motility and no recovery of vision in the left eye. An MRI scan showed a low-intensity lesion which was well delineated from surrounding tissues and resolvable from air (Fig 4). A transcranial approach to the orbit revealed a laceration of the periorbita, and a wooden fragment 2 X 2 X 1 em (Fig 5) was identified and removed from the posterior superior orbit. Intravenous cefazolin was continued for I week postop-
GREEN et al
•
INTRAORBITAL WOOD: RADIOLOGIC DETECTION
Fig 1. Case I. Ocular movements of the left eye are restricted in all directions. Notice proptosis in the left eye and stellate laceration of the left upper eyelid.
Fig 3. Case I. Computed tomographic scan of the left orbit. Sagittal (A) and transverse (B) sections show air (arrows) in superior orbit.
Fig 2. Case I. Orbital tomogram shows air (arrows) in the superior portion of the left orbit.
eratively. The patient did not recover vision in his left eye, and the optic nerve became atrophic. However, his horizontal and vertical ductions improved significantly. An MRI scan 3 weeks after surgery showed no residual foreign body. Case 2. A 25-year-old man was sawing a tree when a branch flew up and struck him nasal to the left medial canthus. A superficialleft upper lid laceration was sutured directly. The patient had increasing left orbital pain and fever and chills over the next 2 days and was referred for further assessment. On ophthalmologic examination at admission his best-corrected visual acuity was 6/6 (20/20) in the right eye and 6/18 (20/60) in the left. Both pupils were equal, round, and reactive to light. Hertel exophthalmometry, with a base of 105 mm, measured 14 mm in the right eye and 22 mm in the left. External examination of the left eye revealed moderate swelling and a purulent discharge but no tract or fistula (Fig 6). Motility was markedly restricted in all fields of gaze. There was marked conjunctival injection and chemosis, a mild iritis, and macular and peripapillary choroidal folds. Applanation tensions were 16 mmHg in the right eye and 33 mmHg in the left. A CT scan showed a linear defect in the orbit which was interpreted as intraconal "air" (Fig 7). Echography identified a large foreign-body structure within the muscle cone. The MRI scan showed a well-delineated low-intensity lesion within the muscle cone inferolaterally (Fig 8). A 2.0 X 1.5 X 0.5 cm wooden fragment was subsequently removed from the orbit (Fig 9). Postoperatively, the patient received a I-week course of intravenous cefazolin. Over the next 7 days, visual acuity in the left
eye returned to 6/9 (20/30), intraocular pressure normalized, and the cellulitis and proptosis resolved. Ocular ductions gradually improved although he still had mild limitations of elevation and abduction at his most recent examination.
DISCUSSION Both patients presented with histories of direct orbital trauma and minor lid lacerations. The injuries appeared superficial, and orbital penetration was not suspected initially. On orbital x-rays and CT scans, the wooden intraorbital foreign bodies were identified as air. As noted in case I, this would be a logical interpretation of the CT findings in the presence of a fracture, In the second case, orbital ultrasound suggested the presence of a foreignbody structure but could not define the type of material. The CT scan showed intraconal air in a linear configuration and no evidence of an orbital fracture. The T 1weighted MRI scans in both cases delineated the foreign body from surrounding tissues, distinguished it from air, and localized it for surgical removal. Review of previous reports suggests that wood, particularly dry wood, is not detected on plain x-rays or CT scans unless it is associated with a radioopaque substance such as metallic paint,S or a foreign-body granuloma has developed. 6 Using a model system, Tate and Cupples7 found that the CT scan (GE 8800) could not detect small pieces of wood. MyllyUi et al 8 reported that the CT scan did not detect intraorbital wood in their two patients. They 609
OPHTHALMOLOGY
•
MAY 1990
•
VOLUME 97
•
NUMBER 5
Fig 5. Case 1. Wood foreign body removed from the left orbit.
Fig 6. Case 2. External photograph shows left proptosis, chemosis, and nasal and upper eyelid lacerations.
Fig 4. Case 1. T,-weighted magnetic resonance imaging scans. Sagittal (A) and transverse (8) sections show a well-delineated foreign body in the superior orbit (arrows).
noted that wood ranged from -618 to +23 Hounsfield Units (units of density relative to an assigned air density value of 0). Thus, wood could easily be mistaken for air, particularly in the presence of fractures and sinus communication. Ultrasound often does not detect orbital wood if it is dry and associated with surrounding air. 10 Reshef et al 4 and Coleman9 reported cases in which they identified intraorbital wooden foreign bodies. The latter authors used A-scan, and Coleman used B-scan ultrasonography. Computed tomographic imaging relies on the differing radiodensities of tissues for their differentiation. The radiodensity of wood is variable and may be similar to that of the orbital tissues, which may account for the potential 610
Fig 7. Case 2. Transverse computed tomographic scan shows air (arrow) in the left orbit and proptosis.
GREEN et al
•
INTRAORBITAL WOOD: RADIOLOGIC DETECIIUN
Fig 9. Case 2. Wood fragment extracted from left orbit.
is a history of injury caused by a pointed object. Second, if orbital fat is present on physical examination indicating violation of the orbital septum, the burden of proof is on the examiner to exclude an intraorbital foreign body. Third, if either orbital x-rays or CT scan reveal "air," particularly in a linear configuration, within the setting of orbital trauma, a high index of suspicion for an organic foreign body should be maintained. In the past, MRI has had a limited role in the evaluation of orbital trauma. Furthermore, it is contraindicated when a metallic foreign body is suspected. 11 However, based on the role MRI scanning played in the management of our two patients, we believe that it should be done in all cases where orbital penetration by a wooden foreign body is suspected and CT scan has not shown a foreign body.
REFERENCES
Fig 8. Case 2. TJ-weighted magnetic resonance imaging scans. Sagittal (A) and transverse (B) sections show a well-delineated foreign body in the left orbit (arrows).
difficulty of recognition. This problem is compounded when air is present; if the wood chip lies on the soft tissueair interface, the diagnosis may be even more difficult. Magnetic resonance imaging depends on the density of protons in the tissue and their relaxation times. These properties of wood are dissimilar enough from those of the soft tissue to allow differentiation. The presence of air also makes interpretation difficult, however. Magnetic resonance imaging has been used in various ways in ophthalmology, 11-13 but its value in detecting intraorbital wood has not previously been reported to our knowledge. A review of both the literature and our cases has three important implications. First, the possibility of an intraorbital foreign body should be suspected in any patient sustaining minor lid trauma, especially when there
1. Ferguson EC III. Deep, wooden foreign bodies of the orbit: a report of two cases. Trans Am Acad Ophthalmol Otolaryngol1970; 74:77887. 2. Macrae JA. Diagnosis and management of a wooden orbital foreign body: case report. Br J Ophthalmol1979; 63:848-51. 3. Brock L, Tanenbaum HL. Retention of wooden foreign bodies in the orbit. Can J Ophthalmol1980; 15:70-2. 4. Reshef OS, Ossoinig KC, Nerad JA. Diagnosis and intraoperative localization of a deep orbital organic foreign body. Orbit 1987; 6:315. 5. Weisman RA, Savino PJ, Schut L, Schatz NJ. Computed tomography in penetrating wounds of the orbit with retained foreign bodies. Arch Otolaryngol 1983; 109:265-8. 6. Grove AS Jr. Computed tomography in the management of orbital trauma. Ophthalmology 1982; 89:433-40. 7. Tate E, Cupples H. Detection of orbital foreign bodies with computed tomography: current limits. AJR 1981; 137:493-5. 8. Myllyla V, Pyhtinen J, Pajvansalo M, et al. CT detection and location of intraorbital foreign bodies. Experiments with wood and glass. ROFO 1987; 146:639-43. 9. Coleman OJ. Reliability of ocular and orbital diagnosis with 8-scan ultrasound. 2. Orbital diagnosis. Am J OphthalmoI1972; 74:704-18. 10. McQuown OS. Ocular and orbital echography. Radiol Clin North Am 1975; 13:523-41. 11. Mafee MF. Magnetic resonance imaging and its simplifications for ophthalmology. In: Reinecke RD, ed. Ophthalmology Annual. New York: Raven Press, 1989; 193-9. 12. Siamovits TL, Gardner TA. Neuroimaging in neuro-ophthalmology. Ophthalmology 1989; 96:555-68. 13. Kelly WM, Paglen PG, Pearson JA. et a1. Ferromagnetism of intraocular foreign body causes unilateral blindness after MR study. AJNR 1986; 7:243-5.
611
Abstract: The authors present two cases in which intraorbital wooden foreign bodies remained undetected after initial ophthalmologic examination and radiologic investigation which included plain orbital x-rays, orbital computed tomography (CT) scans, and, in one case, orbital ultrasound. In each case, subsequent magnetic resonance imaging (MRI) showed a well-delineated low-intensity lesion suggestive of a retained foreign body. Investigation of a case of suspected wooden foreign body in the orbit should include an MRI scan if there is no contraindication, and no foreign body has been defined on CT scan, ultrasound, or plain orbital films. Ophthalmology 1990; 97:608-611
The detection of an intraorbital wooden foreign body is difficult, particularly in cases of apparently minor trauma. Several patients have been reported in whom the usual diagnostic techniques did not detect intraorbital wood. 1-4 In each case, the trauma was described as minor, and the external injury appeared superficial. Orbital xrays rarely detect wooden fragments. 3 Computed tomography (CT) scanning can detect intraorbital wood associated with metallic paint5 or a granuloma,6 but both experimental 7 and clinical 8 studies have shown that CT has little value in detecting dry wood alone. Similarly, orbital ultrasound has had limited success in detecting intraorbital wOOd. 9 ,1O We discuss two patients in whom initial ophthalmologic examination and radiologic investigation were unable to identify intraorbital wood penetration. However, subsequent magnetic resonance imaging (MRI) in each case showed a well-delineated low-intensity intraorbital lesion suggestive of a retained foreign body. In each case a wooden foreign body was surgically removed. To our knowledge, this is the first reported use of MRI to detect intraorbital wood.
Originally received: September 28, 1989. Revision accepted: December 29, 1989. Department of Ophthalmology, The Hospital for Sick Children, Toronto. Department of Ophthalmology, University of Iowa, Iowa City. 3 Department of Radiology, The Hospital for Sick Children, Toronto. 1
2
Reprint requests to Stephen P. Kraft, MD, Department of Ophthalmology, The Hospital for Sick Children, 555 University Avenue, Toronto, Ontario, Canada M5G 1X8.
608
CASE REPORTS Case 1. A 14-year-old boy had been swinging from a tree branch which recoiled and struck his left orbit resulting in immediate total visual loss and complete ptosis. On initial examination 2 hours later, the left upper lid was swollen shut. A central stellate upper lid laceration at the level of the lid crease exposed intact orbicularis fibers with no apparent penetration site, Visual acuity was 6/6 (20/20) in the right eye and no light perception (NLP) in the left eye. Pupil sizes were 4 mm in the right eye and 7 mm in the left. The right pupil reacted briskly to light. The left pupil was unreactive to both direct light and consensual stimulation. The left eye was proptotic. Its motility was severely limited in adduction, elevation, and depression, and partially limited in abduction (Fig I). Forced ductions were positive in all planes, particularly elevation and depression. Ocular examination of the left eye revealed an inferocentral corneal abrasion and moderate flare and cells in the anterior chamber. Funduscopic examination showed moderate Berlin's edema, and the optic disc and vessels were normal. Applanation tensions were 18 mmHg in the right eye and 14 mmHg in the left. Orbital plain films (Fig 2) and CT scans (Fig 3) revealed fractures of the left orbital floor and medial wall and "air" in the superior orbit. The patient was admitted to the hospital and received intravenous cefazolin and methylprednisolone. The upper eyelid wound was explored superficially under local anesthesia, and the laceration was closed. After 3 days, there was no improvement in motility and no recovery of vision in the left eye. An MRI scan showed a low-intensity lesion which was well delineated from surrounding tissues and resolvable from air (Fig 4). A transcranial approach to the orbit revealed a laceration of the periorbita, and a wooden fragment 2 X 2 X 1 em (Fig 5) was identified and removed from the posterior superior orbit. Intravenous cefazolin was continued for I week postop-
GREEN et al
•
INTRAORBITAL WOOD: RADIOLOGIC DETECTION
Fig 1. Case I. Ocular movements of the left eye are restricted in all directions. Notice proptosis in the left eye and stellate laceration of the left upper eyelid.
Fig 3. Case I. Computed tomographic scan of the left orbit. Sagittal (A) and transverse (B) sections show air (arrows) in superior orbit.
Fig 2. Case I. Orbital tomogram shows air (arrows) in the superior portion of the left orbit.
eratively. The patient did not recover vision in his left eye, and the optic nerve became atrophic. However, his horizontal and vertical ductions improved significantly. An MRI scan 3 weeks after surgery showed no residual foreign body. Case 2. A 25-year-old man was sawing a tree when a branch flew up and struck him nasal to the left medial canthus. A superficialleft upper lid laceration was sutured directly. The patient had increasing left orbital pain and fever and chills over the next 2 days and was referred for further assessment. On ophthalmologic examination at admission his best-corrected visual acuity was 6/6 (20/20) in the right eye and 6/18 (20/60) in the left. Both pupils were equal, round, and reactive to light. Hertel exophthalmometry, with a base of 105 mm, measured 14 mm in the right eye and 22 mm in the left. External examination of the left eye revealed moderate swelling and a purulent discharge but no tract or fistula (Fig 6). Motility was markedly restricted in all fields of gaze. There was marked conjunctival injection and chemosis, a mild iritis, and macular and peripapillary choroidal folds. Applanation tensions were 16 mmHg in the right eye and 33 mmHg in the left. A CT scan showed a linear defect in the orbit which was interpreted as intraconal "air" (Fig 7). Echography identified a large foreign-body structure within the muscle cone. The MRI scan showed a well-delineated low-intensity lesion within the muscle cone inferolaterally (Fig 8). A 2.0 X 1.5 X 0.5 cm wooden fragment was subsequently removed from the orbit (Fig 9). Postoperatively, the patient received a I-week course of intravenous cefazolin. Over the next 7 days, visual acuity in the left
eye returned to 6/9 (20/30), intraocular pressure normalized, and the cellulitis and proptosis resolved. Ocular ductions gradually improved although he still had mild limitations of elevation and abduction at his most recent examination.
DISCUSSION Both patients presented with histories of direct orbital trauma and minor lid lacerations. The injuries appeared superficial, and orbital penetration was not suspected initially. On orbital x-rays and CT scans, the wooden intraorbital foreign bodies were identified as air. As noted in case I, this would be a logical interpretation of the CT findings in the presence of a fracture, In the second case, orbital ultrasound suggested the presence of a foreignbody structure but could not define the type of material. The CT scan showed intraconal air in a linear configuration and no evidence of an orbital fracture. The T 1weighted MRI scans in both cases delineated the foreign body from surrounding tissues, distinguished it from air, and localized it for surgical removal. Review of previous reports suggests that wood, particularly dry wood, is not detected on plain x-rays or CT scans unless it is associated with a radioopaque substance such as metallic paint,S or a foreign-body granuloma has developed. 6 Using a model system, Tate and Cupples7 found that the CT scan (GE 8800) could not detect small pieces of wood. MyllyUi et al 8 reported that the CT scan did not detect intraorbital wood in their two patients. They 609
OPHTHALMOLOGY
•
MAY 1990
•
VOLUME 97
•
NUMBER 5
Fig 5. Case 1. Wood foreign body removed from the left orbit.
Fig 6. Case 2. External photograph shows left proptosis, chemosis, and nasal and upper eyelid lacerations.
Fig 4. Case 1. T,-weighted magnetic resonance imaging scans. Sagittal (A) and transverse (8) sections show a well-delineated foreign body in the superior orbit (arrows).
noted that wood ranged from -618 to +23 Hounsfield Units (units of density relative to an assigned air density value of 0). Thus, wood could easily be mistaken for air, particularly in the presence of fractures and sinus communication. Ultrasound often does not detect orbital wood if it is dry and associated with surrounding air. 10 Reshef et al 4 and Coleman9 reported cases in which they identified intraorbital wooden foreign bodies. The latter authors used A-scan, and Coleman used B-scan ultrasonography. Computed tomographic imaging relies on the differing radiodensities of tissues for their differentiation. The radiodensity of wood is variable and may be similar to that of the orbital tissues, which may account for the potential 610
Fig 7. Case 2. Transverse computed tomographic scan shows air (arrow) in the left orbit and proptosis.
GREEN et al
•
INTRAORBITAL WOOD: RADIOLOGIC DETECIIUN
Fig 9. Case 2. Wood fragment extracted from left orbit.
is a history of injury caused by a pointed object. Second, if orbital fat is present on physical examination indicating violation of the orbital septum, the burden of proof is on the examiner to exclude an intraorbital foreign body. Third, if either orbital x-rays or CT scan reveal "air," particularly in a linear configuration, within the setting of orbital trauma, a high index of suspicion for an organic foreign body should be maintained. In the past, MRI has had a limited role in the evaluation of orbital trauma. Furthermore, it is contraindicated when a metallic foreign body is suspected. 11 However, based on the role MRI scanning played in the management of our two patients, we believe that it should be done in all cases where orbital penetration by a wooden foreign body is suspected and CT scan has not shown a foreign body.
REFERENCES
Fig 8. Case 2. TJ-weighted magnetic resonance imaging scans. Sagittal (A) and transverse (B) sections show a well-delineated foreign body in the left orbit (arrows).
difficulty of recognition. This problem is compounded when air is present; if the wood chip lies on the soft tissueair interface, the diagnosis may be even more difficult. Magnetic resonance imaging depends on the density of protons in the tissue and their relaxation times. These properties of wood are dissimilar enough from those of the soft tissue to allow differentiation. The presence of air also makes interpretation difficult, however. Magnetic resonance imaging has been used in various ways in ophthalmology, 11-13 but its value in detecting intraorbital wood has not previously been reported to our knowledge. A review of both the literature and our cases has three important implications. First, the possibility of an intraorbital foreign body should be suspected in any patient sustaining minor lid trauma, especially when there
1. Ferguson EC III. Deep, wooden foreign bodies of the orbit: a report of two cases. Trans Am Acad Ophthalmol Otolaryngol1970; 74:77887. 2. Macrae JA. Diagnosis and management of a wooden orbital foreign body: case report. Br J Ophthalmol1979; 63:848-51. 3. Brock L, Tanenbaum HL. Retention of wooden foreign bodies in the orbit. Can J Ophthalmol1980; 15:70-2. 4. Reshef OS, Ossoinig KC, Nerad JA. Diagnosis and intraoperative localization of a deep orbital organic foreign body. Orbit 1987; 6:315. 5. Weisman RA, Savino PJ, Schut L, Schatz NJ. Computed tomography in penetrating wounds of the orbit with retained foreign bodies. Arch Otolaryngol 1983; 109:265-8. 6. Grove AS Jr. Computed tomography in the management of orbital trauma. Ophthalmology 1982; 89:433-40. 7. Tate E, Cupples H. Detection of orbital foreign bodies with computed tomography: current limits. AJR 1981; 137:493-5. 8. Myllyla V, Pyhtinen J, Pajvansalo M, et al. CT detection and location of intraorbital foreign bodies. Experiments with wood and glass. ROFO 1987; 146:639-43. 9. Coleman OJ. Reliability of ocular and orbital diagnosis with 8-scan ultrasound. 2. Orbital diagnosis. Am J OphthalmoI1972; 74:704-18. 10. McQuown OS. Ocular and orbital echography. Radiol Clin North Am 1975; 13:523-41. 11. Mafee MF. Magnetic resonance imaging and its simplifications for ophthalmology. In: Reinecke RD, ed. Ophthalmology Annual. New York: Raven Press, 1989; 193-9. 12. Siamovits TL, Gardner TA. Neuroimaging in neuro-ophthalmology. Ophthalmology 1989; 96:555-68. 13. Kelly WM, Paglen PG, Pearson JA. et a1. Ferromagnetism of intraocular foreign body causes unilateral blindness after MR study. AJNR 1986; 7:243-5.
611
