Clinical examples of central and lateral erosion of implants: a patient with a spherical implant eroding centrally because of tissue restitution (“cactus syndrome”) (A) and the same patient 2 years later (B), wearing a bridging prosthesis (C); lateral extrusion, as found in hemispheric implants, is probably because of the predominance of the adductive forces of the medial rectus (D-F); and central extrusion (G and H) of a spherical implant, with the patient attempting to use an artificial eye (I).
Surgical technique for implant revision (case shown in Figure 1E): conjunctival edges are reflected clear of the fibrous capsule left after removal of the extruding implant (A); horizontal division of deeper tissues provides access to the posterior part of the socket (B); a polyester-covered acrylic sphere is inserted through a polythene glide, to prevent “cactus syndrome” (C); and the lack of any restitutive forces allows the implant to stay deep in the socket, even before suturing of the superficial tissues (D).
Schematic representation, and avoidance, of “cactus syndrome”: status before implantation of a rough-surfaced sphere, with surface tissues labeled (A-E) (A); forced implantation, dragging superficial tissues into the deeper orbit (B); tissue restitution during the healing phase carries the implant anteriorly, with an apparent “thinning” of tissues overlying the ball (C). Use of a smooth glide prevents dragging of superficial tissues by the implant (D), allowing placement of the ball deep in the socket and closure of the overlying layers without any tension (E).
Sagoo MS, Rose GE. Mechanisms and Treatment of Extruding Intraconal ImplantsSocket Aging and Tissue Restitution (the “Cactus Syndrome”). Arch Ophthalmol. 2007;125(12):1616-1620. doi:10.1001/archopht.125.12.1616
Copyright 2007 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2007
To investigate the initial features and treatment of 26 consecutive patients referred with extruding orbital implants between January 1991 and December 2004.
Retrospective medical record review recording the reason for enucleation, primary implant type, infection when initially seen, time to implant exposure, location of conjunctival defect, and time to surgical revision.
Of the 26 eyes, 16 (62%) were removed after trauma, 3 (12%) because of tumor, 3 (12%) because of infection, and 4 (15%) because of painful blind eyes (percentages do not total 100 because of rounding). Of the 26 eyes, 8 (31%) were right eyes and 15 (58%) were hemispheric implants; 8 implants (31%) were acrylic or glass spheres, and 1 (4%) each was a hydroxyapatite, porous polythene, or bone sphere. Hemisphere extrusion occurred at a mean of 16 years after implantation, significantly later than with spheres (mean, 10 years after implantation; P = .05). The conjunctiva was breached medially in only 1 (sphere) (4%), centrally in 13 (50%), and laterally in 12 (46%). Lateral erosion occurred solely with hemispheres, in contrast to central erosions, in which 10 of 13 (77%) were spheres (P < .001). Twelve patients (46%) underwent surgical revision within a year of extrusion, 7 (27%) within 2 years, and the remaining 7 (27%) at 2 to 21 years.
Exposure of hemispheres occurred later, from pressure erosion at their prominent lateral edge. In contrast, central erosion (in spheres) occurred earlier, because of gradual tissue restitution after forced-ball implantation (“cactus syndrome”). This may be avoided by implantation through a polythene glide.
Intraconal implants are well established in the treatment of anophthalmic sockets,1- 3 and complications include tissue breakdown over the implant. Anterior migration of the implant because of tissue restitution during healing—termed the cactus syndrome by one of us (G.E.R.)—leads to a thinning of surface tissues (Figure 1A), with delayed exposure (Figure 1B) and, later, a frank extrusion (Figure 1G and H) of the implant. Small areas of implant exposure are more common than full extrusion,1,4- 6 although many techniques have been devised to prevent or treat implant exposure.7- 11
We describe a series of patients referred with exposed orbital implants and treated by one of us (G.E.R.). The characteristics of this series suggest a possible mechanism for the late exposure of hemispheric orbital implants.
A retrospective case review was performed for a series of patients referred with extruding orbital implants between January 1991 and December 2004. Information was gathered on the reason for enucleation, type of primary implant, period before symptoms of implant exposure, location of the main conjunctival defect (classified as medial, lateral, or central), and time before surgical revision.
The surgical technique of implant revision is shown in Figure 2. After antiseptic douching of the socket, the conjunctiva is opened at the edge of the defect and the opening is extended medially and laterally across the width of the midline raphe; the upper and lower conjunctival edges are carefully undermined for 3 mm, and the conjunctival flaps are gently slung on 5-0 traction sutures. The exposed implant is removed from its fibrous capsule; in the case of Castroviejo-style implants, this involves tissue incision around the edge “ring” on the anterior face with cruciform division of the union of the recti. The fibrous capsule is inspected, any areas suggestive of epithelial invasion are either excised or ablated with diathermy, and the capsule is divided posteriorly to allow placement of the new spherical implant into the posterior part of the orbit. For primary replacement, the capsular halves and neighboring tissues are slung on 3-0 silk traction sutures and the replacement sphere is implanted deeply in the orbit by passing it through a glide, fashioned from the thumb of a sterile polythene glove. Deep tissues are closed with 5-0 absorbable sutures, conjunctiva is closed with 7-0 sutures, and a large socket conformer is placed after fornical instillation of an antibiotic ointment. The socket is padded firmly for 1 week, and the patient is prescribed a postoperative course of systemic antibiotics.
Clinical case notes were retrieved for 26 patients (20 males [77%]) who were initially seen between the ages of 6 and 62 years (mean age, 41 years; median age, 42 years), this being between less than 1 year and 37 years (mean, 13.5 years; median, 8 years) after the initial surgery undertaken at between the ages of 2 and 58 years (mean age, 27 years; median age, 28 years). Sixteen patients (62%) underwent removal of blind eyes after trauma, 4 (15%) for rubeotic glaucoma, and 3 (12%) each for intraocular tumor or endophthalmitis (percentages do not total 100 because of rounding); 18 (69%) were left eyes (Table). Of the 26 implants, 15 (58%) were hemispheric, 8 (31%) were acrylic or glass spheres (2 with polyester mesh covering), and 1 (4%) each was a hydroxyapatite, porous polyethylene, or bone sphere (percentages do not total 100 because of rounding). One of the hemispheric implants had a gold motility peg.
The patients were initially seen between 1 and 37 years (mean, 16 years) after placement of hemispheric implants, this being significantly different from that of spheres (range, 1-32 years; mean, 10 years) (Mann-Whitney test, P = .05). Overall, the mean age at extrusion was 40.9 years (range, 13-62 years).
Anteromedial implant exposure occurred in only one case—a polyester-covered acrylic sphere that had been malpositioned within the superomedial quadrant of the orbit. Central exposure (13 cases) (Figure 1A, B, G, and H) occurred mainly with ball implants (10 of 13 cases), whereas lateral erosion (12 cases) (Figure 1D-F) occurred solely with hemispheric implants (χ2 = 12.8, P < .001). The extruding implant was removed with simultaneous replacement in 18 patients (69%) and secondary replacement in 7 (27%); 1 patient (4%) elected not to undergo further implantation. Two patients with significant purulent conjunctivitis and markedly inflamed sockets, pretreated with topical and systemic antibiotics, underwent successful primary reimplantation. The surgery was undertaken within a year of developing new symptoms in 12 of 26 patients (46%) and within 2 years of symptoms in 7 patients (27%); the remainder underwent surgery between 2 and 21 years after symptoms occurred.
Postenucleation socket syndrome occurs when there is inadequate volume replacement in the anophthalmic socket, requiring the use of a heavy prosthetic eye that causes lower eyelid stretch, secondary upper eyelid ptosis, enophthalmos, and a deep superior sulcus.12 To avoid postenucleation socket syndrome or complications of a large artificial eye, intraconal implantation is essential for aesthetic rehabilitation after removal of an eye. The ideal volume replacement after removing the eye is a sphere of 21- to 22-mm diameter that, if implanted deep within the orbit, provides a gently convex surface to the socket with about 3-mm central depth, adequate for prosthetic fitting. Several mechanisms have been suggested to explain implant exposure, including infection, edema, hemorrhage, too large an implant, a faulty surgical technique, a poor fitting prosthesis with pressure points, and placement of a motility peg.13- 15 There is also a report16 of an orbital recurrence of retinoblastoma causing implant extrusion. Although these may contribute to exposure, the mechanisms of tissue restitution or long-term adductive rotation of the implant may also play a part in anterior tissue breakdown.
The fact that all laterally eroding implants (Figure 1D-F) were flat fronted, whereas most of the central erosions (Figure 1A, B, G, and H) occurred with spheres, suggests different mechanisms for the 2 types of implant erosion. One of us (G.E.R.) has noted that aged sockets with hemispheric implants tend to have superomedially directed implant faces—the backward tilt having also been shown with imaging studies.17- 19 Backward tilt is due to tissue gravitation within the anophthalmic socket,17,18 whereas medial rotation is (conjecturally) due to a power dominance of the medial rectus over the lateral rectus. Whereas the temporal edge of a hemispheric implant will become more prominent with time, leading to surface pressure erosion over this lateral edge, a sphere will not have a similar time-related change in surface configuration. Logically, flat-fronted intraconal implants should probably be avoided because of this long-term risk of rotation within the aging socket.
Ball implants that are too prominent within a socket—sometimes termed oversized (Figure 3C)—are actually positioned too superficially, this probably resulting from “cactus syndrome.” Cactus syndrome arises when an orbital implant is forced into the tissues during implantation and the implant surface (particularly where rough, as with hydroxyapatite or porous polythene) drags the superficial tissues into the depths of the socket (Figure 3B). The tissues may be closed successfully over the implant but, as with any forced surgical closure, the tension sutures reabsorb and the tissues gradually return to their original relaxed position—a natural restitution of tissues. This restitution is manifest as a progressive migration of the ball toward the surface, with a progressive thinning and eventual breakdown of the attenuated tissues over the implant (Figure 1A, B, and G). The fact that late tissue restitution occurs when a rough-surfaced implant drags fat into the socket depths is analogous to the “spring back” (restitution) that would occur if a cactus was forced into the flocking within a pillow—hence, the term cactus syndrome.
Although many techniques have been described to avoid, or deal with, implant exposure (such as primary revision,20 dermis fat grafting,3 use of a “cap” of sclera21- 24 or other material,25 or patch grafting of exposed implants with temporalis fascia or sclera7,8,10,11), most of these techniques fail to address one of the major causes of implant failure—namely, the superficial positioning of the ball due to late tissue restitution. Indeed, we would suggest that these myriad techniques would become largely redundant if tissue drag is avoided by implantation of all spherical implants, into the intraconal space, using a suitable technique. Retraction of superficial tissues and their careful repositioning may achieve this goal, but we believe deep orbital implantation is easily achieved using the polythene glide.
Correspondence: Geoffrey E. Rose, MS, DSc, FRCS, FRCOphth, Orbital Clinic, Moorfields Eye Hospital, City Road, London EC1V 2PD, England (email@example.com).
Submitted for Publication: June 12, 2007; final revision received July 30, 2007; accepted August 2, 2007.
Financial Disclosure: None reported.