Role of Magnetic Resonance Imaging in the Evaluation of the Hydroxyapatite Orbital Implant

Patrick De Potter, MD, 1 Carol L. Shields, MD, 1 Jerry A Shields, MD,1

Adam E. Flanders, MD,2 Vijay M. Rao, MD2

The role of magnetic resonance imaging (MRI) in the assessment of fibrovascular ingrowth in the integrated hydroxyapatite orbital implant is evaluated. Fifteen patients

who underwent enucleation and placement of a hydroxyapatite orbital implant were evaluated for degree of implant vascularity with gadolinium-OPTA-enhanced MRI with surface coil before drilling the implant. On T1-weighted images, the hydroxyapatite sphere appeared with intermediate signal. After gadolinium-OPTA administration, all patients showed an enhancement in the implant consistent with the presence of fibrovascular ingrowth. The enhancement was most notable in the peripheral portions of the sphere and was seen as early as 5 months after implantation. Comparison of gadolinium­DPTA-enhanced MRI with contrast-enhanced computed tomography, ultrasonography, and color Doppler imaging suggests that these latter techniques are not as helpful in the detection of the fibrovascular tissue in the orbital implant. Bone scan, a technique used by many surgeons, demonstrates fibrovascular ingrowth, but it is limited by its one-dimensional low-resolution image. Because of its three-dimensional capability and its highest resolution, contrast-enhanced MRI with surface coil appears to be the best imaging method for evaluating the hydroxyapatite orbital implant and its fibrovascular ingrowth. Ophthalmology 1992;99:824-830

One of the difficulties with the conventional polymeth­ylmethacrylate orbital sphere after enucleation is the poor motility of the implant/prosthesis complex. This is a result of a lack of transfer of motion from the nonintegrated implant to the prosthesis. To achieve better motility of the prosthesis, a coralline-derived hydroxyapatite sphere was recently introduced as an integrated implant for the

-4anophthalmic socket. 1 The hydroxyapatite implant is

Originally received: September 3, 1991. Revision accepted: October 29, 1991. 1 Oncology Service, Wills Eye Hospital, Philadelphia. 2 Department of Neuroradiology, Thomas Jefferson University Hospital, Philadelphia. Supported by the Macula Foundation, New York, New York, by the Robert Beckman Endowment Fund, Philadelphia, Pennsylvania, and by tbe Eye Tumor Research Foundation Inc, Philadelphia, Pennsylvania. Reprints requests to Jerry A. Shields, MD, Director, Oncology Service, Wills Eye Hospital, Ninth and Walnut Sts, Philadelphia, PA 19107.

directly integrated or attached by a peg to the prosthesis for motility purposes. This implant appears to be well tolerated and has been shown to provide improved pros­thesis motility with very few complications. l-6

The results of clinical and histopathologic studies on the use of hydroxyapatite as a bone substitute in oral, maxillofacial, and orthopedic surgery have been encour­aging.7-11 In a recently reported case, histopathologic study of an orbital hydroxyapatite implant that was removed 4 weeks after implantation showed infiltration by fibrovas­cular tissue within the pores of the sphere. 5 The purpose ofthis study is to investigate the magnetic resonance im­aging (MRI) characteristics of this new implant and to assess the ability of MRI to demonstrate fibrovascular in­growth of the implant, and thus provide information be­fore drilling and peg placement. The advantages of con­trast-enhanced MRI with surface coil are compared with contrast-enhanced computed tomography (CT), ultraso­nography, color Doppler imaging (CDI), and bone scan imaging studies.

824

De Potter et al · MRI Hydroxyapatite

Patients and Methods

There were 120 patients managed on the Oncology Service at Wills Eye Hospital, Philadelphia, who had a hydroxy­apatite orbital implant placed after enucleation between January 1990 and June 1991. Each of these patients had undergone enucleation for management ofan intraocular tumor or blind painful eye. Our selection of patients for the hydroxyapatite implant and the technique of enucle­ation and implantation ofthe hydroxyapatite implant has been described elsewhere.6

We generally use a 20-mm diameter hydroxyapatite sphere for adults and an 18-mm diameter hydroxyapatite sphere for young children. Fresh or banked sclera pre­treated with antibiotics and soaked in 10% povidine-io­dine is prepared and wrapped over the hydroxyapatite implant (Fig 1 ). All six extraocular muscles are attached to their anatomic sites on the scleralized implant (Fig 2). Patients receive their prosthesis 2 weeks after surgery and are re-evaluated 4 to 6 months later. At that visit, they are given the option of having the implant drilled and a peg placed into the implant. This peg, which is then con­nected to the prosthesis, maximizes eventual transfer of motion from the implant to the prosthesis.

Before drilling the implant, orbital integration of the sphere by fibrovascular tissue is necessary to allow toler­ance of the foreign body peg and to minimize the risk of infection. Fibrovascular ingrowth has been demonstrated by others with bone scan.2•4 Because of its known better definition, we chose to use contrast-enhanced MRI to document fibrovascular tissue in the hydroxyapatite sphere in 15 patients. For comparison purposes, CT, A­and B-mode ultrasonography, and CDI were used.

Contrast CT was performed using the General Electric · CT-9800 (Milwaukee, WI). For A- and B-mode ultraso­nography, we used the Ophthascan S Biophysic Medical unit (Clermont Ferrant, France). Color Doppler imaging was performed using the QAD 1 (Quantum Medical Sys­tem, Inc, Issaquah, WA), using a 7.5-MHz linear array transducer. Unlike other surgeons,2•4 we did not perform

Figure L White porous coralline hydroxyapatite implant covered with sclera showing the four scleral windows cut anterior to the equator for rectus muscles reattachment.

Figure 2. The implant is placed in the orbit with the four rectus muscles (arrowheads) sutured to the corresponding anterior lip of the previously designed scleral windows.

bone scan studies because of the lack of three-dimensional visualization and accurate definition. However, represen­tative bone scans were kindly provided by Dr. Arthur C. Perry for comparison with Gd-DPT A-enhanced MRI.

The MRI studies were performed on 1.5 General Elec­tric Signa Unit using tuned, paired 3-inch surface coils. Imaging sequences consisted ofmultiple axial acquisitions, with a 3-mm slice thickness, 1-mm gap, and a field of view of 12 em. Axial, coronal, and sagittal T 1-weighted images were obtained using pulse repetition rate (TR) of 600 msec and echo delay time (TE) of 25 msec. Proton density and T2-weighted images were obtained using TR of2000 msec and TE of20 and 80 msec, respectively. In addition, T 1-weighted images with fat suppression were obtained. A fat suppression technique was used to dimin­ish the usual high signal intensity of orbital fat, thus ex­panding the dynamic range of the image. Gadolinium­DPTA (0.1 mol/kg) was injected intravenously, and the T 1-weighted images, with and without fat suppression, were repeated.

Results

Of the 120 patients with hydroxyapatite implant placed after enucleation on the Oncology Service at Wills Eye Hospital, 16 underwent drilling ofthe ball implant before peg placement. The remaining 104 patients have not had drilling because 40 were ineligible (less than 4 to 6 months from the time of implant placement), 25 were children in whom we have chosen to delay drilling until they are old enough to cooperate for the tedious prosthesis revision, 3 have developed metastasis and/or have died, and 36 were satisfied with their cosmetic appearance and motility and preferred no further treatment. Fifteen patients had a predrilling MRI evaluation and one patient refused MRI studies because of claustrophobia. The specific diagnoses before enucleation in the 15 cases included: uveal malig­nant melanoma in 11 cases, choroidal hemangioma in a blind eye with total retinal detachment in 1 case, and a

825

Ophthalmology Volume 99, Number 5, May 1992

Table 1. Summary of 15 Patients with Hydroxyapatite Orbital Implant Imaged by Gadolinium-DPTA-enhanced Magnetic Resonance Imaging

Patient No./Sex/Eye Diagnosis Pre-enucleation

Implant Size (mm)

Age (yrs) at TimeofMRI

Time Interval (mos) between Enucleation and MRI

1/M/OS Choroidal hemangioma 20 2/M/OD Choroidal melanoma 20 3/M/OS Choroidal melanoma 20 4/F/OD Choroidal melanoma 20 5/F/OS Choroidal melanoma 20 6/M/OD Choroidal melanoma 20 7/M/OD Choroidal melanoma 20 8/F/OS Blind painful eye post-trauma 20 9/F/OD Choroidal melanoma and 20

extraocular extension 10/M/OS Blind painful eye post-trauma 20 11/M/OD Choroidal melanoma 20 12/M/OS Blind painful eye post-trauma 20 13/F/OS Choroidal melanoma 20 14/F/OS Choroidal melanoma 20 15/M/OD Choroidal melanoma 20

MRI =magnetic resonance imaging; OS= left eye; OD =right eye.

blind, painful eye after trauma in 3 cases. The time interval between enucleation and imaging studies ranged from 5 to 13 months (mean, 7 months) (Table 1). The MRI stud­ies were not performed on patients who refused the drilling stage.

The 15 patients (10 males and 5 females) imaged by MRI ranged in age between 16 and 74 years (mean, 46 years). Contrast-enhanced CT was performed on one pa­tient who refused MRI because of claustrophobia. The CT imaging studies were performed 6 months after enu­cleation for choroidal melanoma. A- and B-mode ultrasonography and CDI studies were performed on patient 11.

The hydroxyapatite implant, made of a complex in­ternal porous network filled with fibrovascular tissue and extracellular water, exhibited an intermediate signal, isointense to the opposite globe, on all nonenhanced T,­weighted sequences with or without fat suppression se­quences (Fig 3) (Table 2). The excellent tissue contrast of MRI studies with surface coil allowed for evaluation of the anatomic position of the sphere in the muscle cone with the reattached extraocular muscles. The distal portion of the optic nerve also was seen in all cases.

Gadolinium-DPTA-enhanced T 1-weighted images with or without fat suppression sequences showed en­hancement of the sphere in all patients. The enhancement was diffuse homogeneous in four cases and diffuse non­homogeneous in eight cases. Discontinuous areas of marked hyperintensity, predominantly observed at the periphery of the sphere and suggestive of vascular con­nective tissue, was observed in three cases (Figs 4 and 5). The use of fat suppression techniques highlighted the contrast between subtle differences in signal intensity. In­

38 51/z 74 6 48 5 56 8 34 7 64 6 54 6 59 13 53 8

23 6 50 61fz 16 9 32 51fz 46 12 39 7

creased signal intensity was noted anterior to the equator, corresponding to the site of attachment of extraocular muscles in 11 cases (Fig 6). The proportion ofthe implant that enhanced varied markedly among the patients despite the fact that some patients were imaged after a similar time interval (Table 2). The Trweighted images revealed that the hydroxyapatite ball implant showed hypointense

Figure 3. Nonenhanced axial Tl-weighted images (TR = 600 msec, TE = 20 msec) of a 54-year-old man 6 months after implantation of hydroxy­apatite sphere. The implant (*) in the right orbit appears isointense with respect to the vitreous of the opposite eye. The lateral, medial rectus muscles, and the distal portion of the optic nerve can be clearly identified (P = prosthesis).

826

De Potter et al · MRI Hydroxyapatite

Table 2. Magnetic Resonance Imaging Characteristics in 15 Patients with Hydroxyapatite Orbital Implant

Gd·DPTA T 1•weighted (±fat suppression)

Patient No. T 1•weighted* T2•weighted* Intensity Characteristics of Enhancement

1 lsointense Hypointense +++ Diffuse, nonhomogeneous 2 lsointense Hypointense + Diffuse, nonhomogeneous 3 Isointense NA +++ Diffuse, nonhomogeneous 4 Isointense Hypointense +++ Diffuse, nonhomogeneous 5 lsointense Isointense ++ Diffuse, homogeneous 6 Isointense Hypointense ++ Diffuse, homogeneous 7 !so intense Hypointense ++ Diffuse, nonhomogeneous 8 Isointense Hypointense + Anterior portion, homogeneous 9 Isointense NA ++ Diffuse, nonhomogeneous

10 Isointense Isointense ++ Diffuse, nonhomogeneous 11 Isointense !so intense ++ Diffuse, nonhomogeneous 12 lsointense Hypointense +++ Diffuse, homogeneous 13 Isointense Hypointense ++ Periphery, nonhomogeneous 14 Isointense Hypointense ++ Periphery, homogeneous 15 Isointense NA +++ Diffuse, nonhomogeneous

Gd-DPTA = gadolinium-OPT A; NA = not available due to motion artifact; + = mild enhancement; ++ = moderate enhancement; +++ = marked enhancement. • Signal intensity of the hydroxyapatite orbital implant with respect to the vitreous of the opposite eye.

or isointense signal with respect to the vitreous of the Because of the sound attenuation created by the calcific opposite eye. The hydroxyapatite implant was not clearly hydroxyapatite material, the presence offibrovascular in­evaluated on T 2-weighted images in 3 cases because of growth in the sphere could not be detected either by A­motion artifact. and B-mode ultrasonography or by CDI (Fig 8). However,

On noncontrast CT studies, the integrated hydroxy­apatite implant appeared homogeneous and as dense as the mature bone of the orbit. The CT contrast agents did not improve the detection of fibrovascular ingrowth in the sphere 6 months after implantation because of over­riding density of the hydroxyapatite sphere (Fig 7).

Figure 5. Nonenhanced (above) and gadolinium-OPT A-enhanced (below) axial Trweighted images with fat suppression technique (TR = 600 msec, TE = 20 msec) of a 56-year-old woman 8 months after implantation of hydroxyapatite sphere (*) in the right orbit. Marked nonhomogeneous

Figure 4. Gadolinium-DPTA enhanced axial T 1-weighted image of the enhancement of the ball suggests fibrovascular ingrowth (P = prosthesis). same patient in Figure 3 showing the diffuse and nonhomogeneous en­ Published courtesy of Archives of Ophthalmology 1992; 110:333-8. Copy­hancement of the hydroxyapatite sphere. right 1992 American Medical Association.

827

Ophthalmology Volume 99, Number 5, May 1992

Figure 6. Nonenhanced (above) and gadolinium-DPTA-enhanced (below) axial T 1-weighted images (TR = 600 msec, TE = 20 msec) of a 64-year­old man with hydroxyapatite implant placed 6 months before MRI studies. After gadolinium administration, scattered hyperintensity signals (arrow­heads) are noted in the ball at the insertion of both horizontal rectus muscles, suggesting fibrovascular ingrowth through the scleral windows (P = prosthesis).

COl showed detectable vascularity at the insertion of ex­traocular muscles attached to the hydroxyapatite implant.

Discussion

Porous hydroxyapatite is a material derived from the skeletal structure of specific marine corals composed pri­marily of calcium phosphate with a regular system of in-

Figure 7. Enhanced axial CT image of a 43-year-old man. The hydroxy­apatite implant (*) in the left orbit appears as dense as the mature orbital bone. The presence of fibrovascular tissue can not be identified (P = prosthesis).

Figure 8. Color Doppler imaging of a hydroxyapatite ball implant (right eye) 9 months after insertion. The sound attenuation by the calcium phosphate component of the sphere (small white arrows) does not allow detection of blood flow. Notice the blood flow at the muscle insertion (black arrow). Color Doppler imaging of the opposite normal eye (left eye) shows the blood flow in the optic nerve and central retinal vessels (large white arrow).

terconnecting pores of approximately 500-,um diame­ter.7·12 Because it has been shown to be nontoxic, biocom­patible, and nonallergenic in humans,7 the porous hydroxyapatite implant has gained tremendous popularity as bone graft substitute over the past 15 years.7·10- 12 Clin­ical applications of hydroxyapatite, because of its bio­mechanical function, have included periodontal defects grafts, 10 post-traumatic long bone grafts (Bucholz, un­published data, presented at AAOS Annual Meeting, 1986), and maxillofacial skeleton augmentationY Be­cause of the lack ofblind passages or cui-de-sacs that could sequester organisms and predispose to infection, the po­rous material acts as a scaffold for the rapid ingrowth of

9 10 14vascularized connective tissue and bone.7· • ·The major difference of the orbital hydroxyapatite im­

plant from the maxillofacial and orthopedic implant is that the material is not in direct contact with bone in the orbit. Shields and associates5 demonstrated fibrovascular ingrowth in an orbital hydroxyapatite sphere implanted 4 weeks earlier. The fibrovascularity was most extensive near the precut scleral windows that were designed for rectus muscle attachment. It was demonstrated that fi­brovascular tissue extended approximately 3 mm centrally at the 4-week visit. These findings confirmed the theory that the hydroxyapatite implant is vascularized when it is placed in soft tissue such as the human orbit.

The three major theoretical advantages of orbital fi­brovascularized implant are that 1) it is less likely to be­come infected because the implant is vascularized with host blood vessels so that access to host immune defenses are improved, 2) it is less likely to extrude because it is biologically fixed in the orbital soft tissues, and 3) it tol­erates the foreign body peg. 1Shields and associates3 and Dutton4recently reported no cases of orbital infection or orbital extrusion in their series of orbital hydroxyapatite implants. The position of the hydroxyapatite sphere an­

828

De Potter et al · MRI Hydroxyapatite

chored via the extraocular muscles to orbital tissue within the muscle cone explains the absence of extrusions. 3- 5

The application ofimaging techniques in the evaluation of conventional polymethylmethacrylate orbital implant has been limited. Computed tomography has shown utility in assessing implant migration, 15 orbital inflammation, or orbital recurrence of intraocular malignant tumors. 3

However, because of the biophysical properties of the cal­cium phosphate hydroxyapatite material, this integrated orbital sphere appears as dense as the mature bone on CT scans. Computed tomography with contrast agents did not detect fibrovascular ingrowth in the hydroxyapatite ball in our case because of the expected over-riding high density of the calcified hydroxyapatite material.

Real time A- and B-mode ultrasonography and more recently CDI have improved diagnostic imaging in eval­uating intraocular and orbitallesions.6•16 Color Doppler imaging allows for noninvasive ultrasonographically guided qualitative assessment of blood flow. 16 Because of the dramatic sound attenuation created by the calcific hydroxyapatite material, the presence ofblood flow could not be detected either by A- and B-mode ultrasonography or by CDI.

In some centers, radionuclide bone scanning with in­travenous injection of technetium 99m is used to assess fibrovascular ingrowth in hydroxyapatite orbital implant 6 months after insertion.2.4 The presence ofgross increased activity at the site of the implanted orbit is interpreted as vascularization of the hydroxyapatite sphere (Fig 9). However, the limited tissue specificity of the bone scan image does not allow high-resolution visualization of the intraorbital structures and the integrated hydroxyapatite sphere. Moreover, the absence ofa three-dimensional im­age lacks discrete localization of the area of the increased uptake. It has not been clearly delineated whether the orbital hyperactivity area represents fibrovascular in­growth or simply vascular peri-implant scar tissue after surgery.

Magnetic resonance imaging with surface coil is a su­perior imaging tool in the field of ophthalmology. 17 Fat suppression techniques on T 1-weighted images provide

Figure 9. Technetium 99m bone imaging of the head and neck of a patient implanted with a hydroxyapatite sphere 6 months previously. There is diffuse nonspecific high uptake in the right orbit on the anterior and right lateral projections. (Courtesy of Dr. Arthur C. Perry, San Diego, CA.)

high-contrast evaluation ofthe orbit and exploit subtle areas ofenhancement. 18 The paramagnetic gadolinium+3 (Gd+3)

chelated to DPTA additionally enhances areas ofrelatively slow-flowing blood and increased vascularity. 19

-23

The calcific hydroxyapatite implant before implanta­tion in the orbit is filled with air and is of low signal in­tensity on all pulse sequences. After implantation in the orbit, however, the implant gradually becomes filled with blood vessels.5 We found that the implant at this stage exhibited an intermediate signal on all T1-weighted se­quences. The intensity ofthe signal appeared slightly more hyperintense on fat suppression sequences because of the change in dynamic range. The T 1-weighted findings were attributed to the filling of the porous channels with fluid and vascular connective tissue.5 After intravenous ad­ministration of gadolinium-OPTA on T1-weighted im­ages, variable hyperintense signal areas in the hydroxy­apatite implant suggested again the presence of fibrovas­cular ingrowth. Gd-DPTA-enhanced T 1-weighted images differentiated fibrovascular tissue from simple fluid con­tained within the pores. Because of the bright signal of the surrounding orbital fat on T 1-weighted images, subtle areas of enhancement were less conspicuous. Nonen­hanced and Gd-DPTA-enhanced T 1-weighted images with fat suppression improved subtle contrast differences in signal intensity by expanding the dynamic range of the image. 18

With T2-weighted images, the presence ofextracellular water in the hydroxyapatite implant appeared as areas of hyperintensity, similar to normal vitreous in the contra­lateral globe depending on the extracellular water. How­ever, interference from patient motion artifacts inherent to the length ofT2 sequences makes the interpretation of T rweighted images difficult.

There was no definite correlation between the extent ofhydroxyapatite implant enhancement and the time in­terval that elapsed after implantation. This observation may reflect individual variations in angiogenesis. The fi­brovascular tissue seemed to originate at the sites of the muscle insertions. The presence of enhancement rather than the characteristics of this enhancement were our guide for drilling. No signs of inflammatory or recurrent neoplastic process were noted.

The hydroxyapatite ball implant appears to offer a sig­nificant improvement of prosthesis motility with excellent tolerance and very few complications. Because ofits high­est definition, MRI with surface coil is the best imaging technique for evaluating the hydroxyapatite orbital im­plant for vascularization before drilling of the sphere and peg placement. Although MRI studies are somewhat more expensive than other imaging methods, we believe that, based on our limited experience, obtaining Gd-DPTA­enhanced T 1-weighted images with or without fat suppression is the procedure of choice in assessing fi.bro­vascular ingrowth in the hydroxyapatite orbital implant. Comparative studies with bone scan and contrast-en­hanced MRI warrant further research. Future imaging studies before 5 months after enucleation may indicate that earlier drilling of the implant may be possible.

829

Ophthalmology Volume 99, Number 5, May 1992

References 13. Salyer KE, Hall CD. Porous hydroxyapatite as an onlay bone-graft substitute for maxillofacial surgery. Plast Reconstr

1. Perry A C. Advances in enucleation. Ophthalmol Clin North Am 1991 ;4: 173-82.

2. Perry A C. Integrated orbital implants. Adv Ophthalmic Plast Reconstr Surg 1990;8:75-81.

3. Shields CL, Shields JA, De Potter P. Hydroxyapatite orbital implant after enucleation. Experience with initial 100 con­secutive cases. Arch Ophthalmol [in press].

4. Dutton JJ. Coralline hydroxyapatite as an ocular implant. Ophthalmology 1991 ;98:370-7.

5. Shields CL, Shields JA, Eagle RC Jr, De Potter P. Histo­pathologic evidence of fibrovascular ingrowth four weeks after placement of the hydroxyapatite orbital implant. Am J Ophthalmol 1991;111:363-6.

6. Shields JA, Shields CL. Intraocular Tumors. A Text and Atlas. Philadelphia: WB Saunders Co, 1992. Chap 3.

7. Piecuch JF, Topazian RG, Skoly S, Wolfe S. Experimental ridge augmentation with porous hydroxyapatite implants. J Dent Res 1983;62:148-54.

8. White E, Shors EC. Biomaterial aspects of Interpore-200 porous hydroxyapatite. Dent Clin North Am 1986;30:49­67.

9. Sartoris DJ, Gershuni DH, Akeson WH, et al. Coralline hydroxyapatite bone graft substitutes: preliminary report of radiographic evaluation. Radiology 1986;159:133-7.

10. Kenney EB, Lekovic V, Sa Ferreira JC, et al. Bone formation within porous hydroxylapatite implants in human perio­dontal defects. J Periodontal 1986;57:76-83.

11. Rosen HM. Surgical correction of the vertically deficient chin. Plast Reconstr Surg 1988;82:247-56.

12. Piecuch JF. Extraskeletal implantation of a porous hy­droxyapatite ceramic. J Dent Res 1982;61:1458-60.

Surg 1989;84:236-44. 14. Holmes RE. Bone regeneration within a coralline hydroxy­

apatite implant. Plast Reconstr Surg 1979;63:626-33. 15. Smit TJ, Koornneef L, Zonneveld FW, et al. Primary and

secondary implants in the anophthalmic orbit. Preoperative and postoperative computed tomographic appearance. Ophthalmology 1991 ;98: 106-10.

16. Lieb WE, Shields JA, Cohen SM, et al. Color Doppler im­aging in the management of intraocular tumors. Ophthal­mology 1990;97: 1660-4.

17. Bilaniuk LT, Atlas SW, Zimmerman RA. Magnetic reso­nance imaging ofthe orbit. Radiol Clin North Am 1987;25: 509-28.

18. Atlas SW, Grossman RI, Hackney DB, et al. STIR MR imaging of the orbit. AJNR 1988;9:969-77.

19. Healy ME, Hesselink JR, Press GA, Middleton MS. In­creased detection ofintracranial metastases with intravenous Gd-DPTA. Radiology 1987;165:619-24.

20. Kuharik MA, Edwards MK, Farlow MR, et al. Gd-enhanced MR imaging of acute and chronic experimental demyelin­ating lesions. AJNR 1988;9:643-8.

21. Zimmerman CF, Schatz NJ, Glaser JS. Magnetic resonance imaging of optic nerve meningiomas: enhancement with gadolinium-DPTA. Ophthalmology 1990;97:585-91.

22. Goldstein EJ, Burnett KR, Hansell JR, et al. Gadolinium DPTA (an NMR proton imaging contrast agent): chemical structure, paramagnetic properties and pharmacokinetics. Physiol Chern Phys Med NMR 1984;16:97-104.

23. Haughton VM, Rimm AA, Czervionke LF, et al. Sensitivity of Gd-DPTA-enhanced MR imaging of benign extraaxial tumors. Radiology 1988;166:829-33.

830