Secondary Logo

Journal Logo

Original Investigations

Volumetric Versus Empirical Determination of Enucleation Implant Size

Custer, Philip L. M.D., F.A.C.S.*; Maamari, Robi N. M.D.*; Huecker, Julia B. M.S.*; Gordon, Mae O. Ph.D.*,†

Author Information
Ophthalmic Plastic and Reconstructive Surgery: September/October 2021 - Volume 37 - Issue 5 - p 414-419
doi: 10.1097/IOP.0000000000001884
  • Free

Orbital implant size is one of the primary variables the surgeon can modify during the enucleation procedure. Historically, ophthalmologists used standard sized implants in all patients, occasionally adjusting for patient age or the depth at which the implant was placed in the orbit.1–3 Because of concerns for wound breakdown and implant exposure, smaller implants were often inserted.4,5 However, small implants can contribute to the development of anophthalmic socket syndrome with a deep superior sulcus. Several authors have advocated individualizing implant size, hoping to maximally replace the volume lost during the enucleation procedure.6–9 Suggested techniques include the use of preoperative ultrasound, intraoperative volume measurements, or empirically with sizing implants.10–13 In addition to volume, there may be secondary factors that should be considered when selecting implant size, including orbital tissue fibrosis or atrophy from previous disease or treatment.

The purpose of this project was to compare 2 techniques for implant sizing: empirically using sizing implants and intraoperative volume measurements of enucleated tissue. The effects of preoperative and surgical findings on implant size were also investigated.


This Health Insurance Portability and Accountability Act-compliant, observational cohort study was approved by the Institutional Review Board at Washington University School of Medicine and adhered to the ethical principles outlined in the Declaration of Helsinki. A retrospective chart review was performed on adult patients undergoing enucleation from January 1, 2007 to December 31, 2016. Individuals with unavailable medical records were excluded, as were those with preoperative conditions or interventions potentially affecting orbital function or obtaining accurate measurements. Patient demographics, intraoperative observations, surgical technique, implant sizing method, and clinical measurements were reviewed. Reference photographs (Fig. 1) were used to classify the depth of the superior sulci (none, mild, moderate, and marked).11,14,15 Postoperative measurements and photographs were obtained a minimum of 2 months after surgery. Statistical analysis was performed with correlation coefficients, chi-square, and t-tests using SAS (V9.4). Means were compared with t-tests and nonparametric methods, with similar results.

FIG. 1.
FIG. 1.:
Reference photographs used to classify superior sulcus depth. A, No increase in superior sulcus depth. B, Mild increase. C, Moderate increase. D, Marked increase.

Enucleation Procedure.

All patients underwent enucleation. After globe removal, the implant size was determined either by volume displacement or with sizing implants. The volume method was preferentially used in those individuals with apparent preoperative globe symmetry. The size of the eye and any attached enucleated tissue was estimated during surgery through displacement of saline in a sterile graduated cylinder (“volume group”).8 The maximal implant volume was calculated by subtracting the minimum volume of a prosthetic (2 ml) and volume of donor sclera (average: 0.9 ml). Alternatively, implant size was empirically determined by temporarily inserting both an implant and conformer, realizing the typical conformer is thinner than a prosthesis (“empirical group”). With the goal of matching the appearance of the opposite eye, larger or smaller implants were tested if the size seemed inappropriate. All implants were covered with donor sclera, which was fixated with 5-0 polyglactin sutures. The implants were placed in the intraconal space, and the 4 rectus muscles were attached to the implant with 6-0 polyglactin sutures at positions slightly anterior to their natural attachment sites. One or both oblique muscles were also fixated in most patients. Anterior Tenon fascia was usually approximated in 2 layers using 6-0 polyglactin sutures with deep interrupted closure and a more superficial running closure. Single-layer closure, or mobilization and separate closure of posterior Tenon fascia, was performed in patients with attenuated tissue. Conjunctiva was sutured with 7-0 chromic. Conformer size was determined during surgery by the depth of the fornices and the ability to easily close the lids over the conformer. Medial and lateral temporary 6-0 chromic suture tarsorrhaphies were frequently placed.


Among the 223 enucleations performed during the study period, 84 patients were excluded, most commonly for: unavailable medical records (n = 24), primary dermal fat graft (n = 18), orbital trauma (n = 9), pediatric patient (n = 9), bilateral visual loss, or surgery (n = 8). A single patient developed implant exposure within 1 month after primary enucleation for choroidal melanoma, treated with initial wound repair, followed by implant replacement with the same sized sphere 5 weeks after surgery. Information on the remaining 139 study patients is shown in Table 1. Members of the empirical group were on average younger, had a greater incidence of preoperative relative enophthalmos, and more frequently had conjunctival-Tenon fascial fibrosis encountered during surgery. A history of intraocular malignancy was more common in the volume group. Thirty-four patients were only included in the preoperative and surgical analysis because of follow up <2 months (n = 21) or inadequate postoperative data (n = 13). Average follow up in the remaining 105 patients was 33.1 months (range: 2.3–87.0 months).

- Preoperative demographics, intraoperative findings, and postoperative results for the 2 implant sizing methods
Total Volume group Empirical group P
Number patients N = 139Male: 80; Female: 59 39.6% (55/139)Male: 27; Female: 28 60.4% (84/139)Male: 53; Female: 31
Age in years: mean (range) 59.7 (18.6–91.6) 63.1 (27.0–90.8) 57.4 (18.6–91.6) 0.0564
Intraocular malignancy* 32.4% (45/139) 60.0% (33/55) 14.3% (12/84) <0.0001
Preoperative enophthalmos (≥2 mm) 31.4% (32/102) 6.4% (3/47) 52.7% (29/55) <0.0001
Intraoperative fibrosis 45.5% (61/134) 25.0% (13/52) 58.5% (48/82) 0.0010
Implant size: mean, SD, group size 21.0, SD = 1.08, n = 139 21.4, SD = 0.93, n = 55 20.8, SD = 1.12, n = 84 0.0016
Large conformer 80.6% (112/139) 90.9% (50/55) 74.7% (62/83) 0.0171
Postoperative findings§
Implant enophthalmos (mm): mean, SD, group size 7.1, SD = 1.62, n = 62 7.2, SD = 1.46, n = 26 7.0, SD = 1.76, n = 36 0.6562
Prosthetic enophthalmos (mm): mean, SD, group size 1.6, SD = 1.33, n = 76 1.8, SD = 1.08, n = 32 1.4, SD = 1.5, n = 44 0.1990
Differences in sample sizes represents availability of data in each category.
*Remainder of patients had blind, painful, disfigured, or phthisical eyes.
Five patients omitted for intraoperative conjunctival resection.
Medium: n = 20, small: n = 4, none: n = 2, and not noted: n = 1.
§Among 105 patients with postoperative data.

Spherical implants of the following materials were used: acrylic (124/139), porous polyethylene (14/139), and hydroxyapatite (1/139). Large conformers were more commonly placed in the volume group than in empirical patients. Patients with intraoperative fibrosis were less likely to receive a large conformer (62%, 37/60) than those without (95%, 70/73, p < 0.001). Large conformers were used in 87.0% (67/77) of patients with implants ≥22 mm, compared with 73.8% (45/61) of those with smaller implants (p = 0.14).

Figure 2 shows the distribution of implant sizes for the 2 methods. Odd sized and implants larger than 22 mm were only used during the early stage of the study. They were not considered in the majority of patients, as the surgeon transitioned to primarily using acrylic implants that were available in limited sizes. Average implant size was greater in males (21.3 mm, SD = 0.99, n = 80) than in females (20.7 mm, SD = 1.13, n = 59) (p = 0.0027), and also in the volume group compared with empirical cases. Spheres ≥ 22 mm diameter were placed in 46% (26/61) of patients with, and 63% (46/73) without intraoperative fibrosis (p = 0.047). There were 16 patients with attenuated Tenon fascia due to involutional, pathologic, or prior surgical changes, in whom closure was performed in a single-layer fashion or through mobilization of posterior Tenon fascial flaps to allow two-layer closure. These patients more frequently received implants ≥ 22 mm (68.8%, 11/16) when compared with those in which standard closure (48.8%, 60/123) was performed (p = 0.1328).

FIG. 2.
FIG. 2.:
Distribution of implant sizes between the empirical and volumetric methods.

The Hertel exophthalmometry base (“base”) was measured with the instrument placed snugly overlying the lateral orbital rim at each canthus. The mean base was greater in patients receiving 22 mm implants (98.4 mm, SD = 4.47, n = 62) compared to 20 mm implants (95.2 mm, SD = 3.82, n = 48) (p = 0.0001). Spheres larger than 20 mm were inserted in 80.4% (41/51) of males with a base >95 mm, but in only 37.5% (6/16) with a narrower base. Implants larger than 20 mm were placed in 76.9% (10/13) of females with a base of >97 mm, but in only 30% (12/40) with a narrower base.

The mean contralateral eye exophthalmometry measurement was greater in males (19.6 mm, SD = 2.91, n = 67) than females (18.4 mm, SD = 3.04, n = 53, p = 0.0364), and in patients with 22 mm (19.5 mm, SD = 3.0, n = 62) versus 20 mm implants (18.3 mm, SD = 3, n = 49, p = 0.042). When considering both the contralateral reading and base measurement, implants larger than 20 mm were inserted in 88.9% (8/9) of females with a Hertel base of >97 mm and contralateral >18 mm, and 92.3% (24/26) of males with a base >95 mm and contralateral >19 mm.

The postoperative position of the implant, as measured by Hertel, was unrelated to implant size (r = 0.084, p = 0.517, n = 62). The degree of relative implant enophthalmos was calculated by subtracting the postoperative exophthalmometry reading of the implant from that of the contralateral eye (Fig. 3).6,8 There was no difference in the mean implant enophthalmos between the empirical and volume groups (p = 0.6562), between patients with or without fibrosis (p = 0.6049), or among patients with 20 mm (6.9 mm, 1.8 SD, n = 28) versus 22 mm (7.1 mm, 1.6 SD, n = 26) implants (p = 0.6602). Of the patients with postoperative measurements, 21.0% (13/62) had more than 8 mm of implant enophthalmos, suggesting an undersized implant. The incidence of such excessive enophthalmos was similar between the 2 methods (empirical: 8/36; volume: 5/26; p = 0.7752). A 22-mm sphere had been placed in 7 of these cases.

FIG. 3.
FIG. 3.:
The amount of relative implant enophthalmos as determined by exophthalmometry for the 2 implant sizing methods.

Implant size was not correlated with prosthetic exophthalmometry readings (p = 0.6086). Relative prosthetic enophthalmos was calculated by subtracting the postoperative measurement of the prosthesis from that of the contralateral eye. There was no difference in the average between empirical and volume methods (p = 0.1990), or between patients with or without fibrosis (p = 0.6541).

The depth of the superior sulcus was graded on photographs. Before surgery, operative side exophthalmometry was progressively lower with greater sulcus depth (p = 0.0007). A very deep preoperative sulcus was more common in patients subsequently found to have intraoperative fibrosis (p = 0.0066) or to have placement of a smaller implant (p = 0.010). Relative sulcus depth was determined by comparing the appearance of the contralateral eye. Preoperatively, those patients whose enucleation side sulcus was moderately markedly deeper had lower mean exophthalmometry (p < 0.0001) compared with those with normal symmetrical sulci.

There was no significant difference (p = 0.6394) in the frequency of relative postoperative sulcus deformity between the sizing methods (Fig. 4). Looking at combined data from the 2 groups, the anophthalmic sulcus depth remained stable in 45.4% (44/97), improved in 11.3% (11/97), and worsened in 43.3% (42/97) of patients after surgery. The sulcus became relatively deeper in 23.1% (9/39) of patients with a 20 mm and 50% (24/48) with 22 mm implants (p = 0.0101). Exophthalmometry measurements of the implant (p = 0.0004) and prosthetic (p = 0.0014) progressively decreased with increasing depth of the ipsilateral sulcus.

FIG. 4.
FIG. 4.:
Relative postoperative depth of the anophthalmic superior sulcus compared with the contralateral side for the 2 implant sizing methods.


Individualizing enucleation implant size has been advocated to reduce volume deficit and the development of the anophthalmic, or postenucleation socket syndrome, which is characterized by a deep superior sulcus, ptosis, lower eyelid laxity, and less commonly upper eyelid retraction.13,15–18 Previously, these changes were partially attributed to postoperative atrophy of orbital fat.19 Currently, it is believed this condition is primarily related to volume loss from resection of the globe, with subsequent redistribution of orbital soft tissue.20 Enucleated tissue volume is preferably replaced by an appropriately sized implant as opposed to a larger prosthetic.7,21,22 The ideal prosthetic has a volume of 2–3 ml and thickness of at least 4 mm.9,20,22 Larger prostheses may have decreased movement and are associated with lower eyelid ectropion.22 Thaller measured enucleated eyes by volume displacement, recommending that larger implants be inserted.9 A similar study found a range in the volume of “normal” appearing enucleated globes, showing that individualizing implant size resulted in less postoperative implant “enophthalmos.”8 Kaltreider et al.7 used contralateral ultrasound in anophthalmic patients to retrospectively calculate the volume needing replacement, finding that 76% of adult patients could have accommodated a bigger sphere than originally inserted. Using A-scan in a prospective study, this same lead author inserted implants greater than 22 mm in 18.5% of patients undergoing enucleation, evisceration, or secondary reconstruction.6 In addition to reducing volume deficit, larger enucleation implants provide better horizontal and vertical implant movement, which potentially can be transmitted to the prosthesis.23 Postoperative levator function is also greater in patients who have received larger implants.24 Reflecting the trend toward the use of bigger enucleation implants, spheres of more than 20 mm in diameter were placed in 56.1% of the patients in the present study. Nevertheless, we estimate that 21% of the patients could have accommodated an implant larger than inserted, and at least 11% could have received a sphere of more than 22 mm diameter.

The low correlation between implant size and position as measured by exophthalmometry may seem counterintuitive. This poor correlation may in fact validate that sizing methods are effective in addressing the volume needs of individual patients, who have variation in the size of their globes and bony orbits. A smaller implant in a shallow or crowded orbit could yield a similar exophthalmometry reading as a larger implant in a spacious orbit. Implant position may also in part be determined by the attachment sites of the extraocular muscles, which were consistent throughout the study.

In this project, the authors compared the results of 2 techniques for individualizing implant size: empirically using sizing implants versus volumetric displacement of enucleated tissue.8,12,13,21 The patient populations in the 2 groups differed. The empirical method was preferentially used in patients with phthisis, while volume displacement was utilized in those with symmetrically appearing globes. Even though the empirical group was more likely to have preoperative enophthalmos, after surgery there was not a significant difference in the amount of relative implant or prosthetic enophthalmos between the 2 techniques. The 2 methods had a similar incidence of a deeper anophthalmic sulcus. Compared with a series where preoperative ultrasound was used to determine implant size, the present study had a higher incidence of prosthetic enophthalmos and deep superior sulcus.6 This difference is likely explained by the frequent use of implants larger than 22 mm in the ultrasound project. Postoperative deepening of the superior sulcus was more common in our patients who received larger (22 mm) implants, indicating that many of these individuals may have accommodated a bigger sphere.

Implant size is calculated in the volume displacement method, while using sizing implants forces the surgeon to make an empirical decision during surgery. Tyers and Collin13 advocated placing the biggest implant that would fit comfortably in the orbit. Verhoekx attempting to insert the largest implant that would allow wound closure without tension, placing 20 mm implants in 46.5% and 22 mm implants in 41.3% of enucleations.12 These results are very similar to those in our empirical group (20 mm: 48.8%; 22 mm: 41.7%). Presumably, the empirical method of implant sizing is easily reproducible by other surgeons.

An anatomic study using CT scans found variation in globe dimensions among healthy adults, with the transverse diameter of eyes being highly correlated to orbital width.25 Analysis of MRI scans also found that eyeball and orbital volume were positively associated.26 While orbital size was not directly assessed in the present study, exophthalmometry base may be an indirect measurement. Males with a base >95 mm, and females >97 mm, more commonly received spheres larger than 20 mm diameter. Smaller implants were frequently used when the base was narrower. Greater contralateral exophthalmometry readings were also associated with placement of larger implants. Thus, preoperative exophthalmometry measurements may be useful in guiding the surgeon to choose the appropriate implant size.

Nunery et al.27 reported that larger implants were not associated with an increased risk of exposure in patients who received either silicone or hydroxyapatite spheres. Kaltreider and Newman28 also found that exposure risk was independent from implant size. A large review of postenucleation complications discovered no relationship to implant size.21 Conformer size was not limited by larger implants when using the techniques the authors describe, suggesting that bigger implants do not compromise the depth of the fornices. Many of the patients had complex ocular histories and intraoperative fibrosis of conjunctiva-Tenon fascia, conditions that potentially could increase the risk of wound breakdown or limit implant size.29 The authors use primary dermal fat grafts when it appears that closing conjunctiva will result in significant forniceal shortening. While mean implant size in patients with intraoperative fibrosis was slightly reduced, spheres of at least 22 mm were inserted in 46% of patients of these cases. Intraoperative scarring did not predispose patients to an orbital volume deficit as determined by implant and prosthetic exophthalmometry. Likewise, attenuated Tenon fascia requiring nonstandard closure did not limit implant size. During the study period there was one excluded case of early implant exposure following primary enucleation for melanoma. There were no other incidences of wound breakdown or exposure extrusion. These results suggest that when using the described surgical techniques, preexisting attenuated or fibrotic conjunctiva-Tenon fascia may not increase the incidence of implant exposure.

The retrospective nature of this project predisposes to several potential limitations. Not all patients in the cohort had complete data. However, clinical information was available in the majority of cases, limiting the potential for selective bias. The patient populations in the 2 groups were different. The volume displacement method was primarily utilized in patients with “normal sized” globes. This technique is not easily used when the globe is phthisical, while the empirical method is applicable in all cases. The authors emphasize that the results are a reflection of the enucleation technique described, with an attempt to maximally replace volume with larger implants. Surgeons who are hesitant to insert large spheres or who manage the extraocular muscles differently may not achieve similar outcomes. Pediatric patients and the results of evisceration or secondary reconstruction were not evaluated in this project.

In summary, previous studies have demonstrated the advantages of individualizing enucleation implant size in reducing postoperative volume deficit.6–8 While implants should not be intentionally oversized, larger spheres provide improved function without increased risk of complications.21,24,30 Determining enucleation implant size either empirically or by volume displacement achieved similar results, as measured by depth of the anophthalmic superior sulcus, and degree of relative implant or prosthetic enophthalmos. When using the empirical sizing method, preoperative exophthalmometry measurements can help guide the surgeon as to what implant size to initially consider. Greater availability of odd sized and acrylic implants bigger than 22 mm would be helpful. Additionally, the intraoperative presence of conjunctival-Tenon fascial fibrosis often does not limit the ability to insert larger implants, or predispose patients to wound breakdown and implant exposure.


1. Shields CL, Shields JA, De Potter P, et al. Problems with the hydroxyapatite orbital implant: experience with 250 consecutive cases. Br J Ophthalmol 1994;78:702–706.
2. Nunery WR, Cepela MA, Heinz GW, et al. Extrusion rate of silicone spherical anophthalmic socket implants. Ophthalmic Plast Reconstr Surg 1993;9:90–95.
3. Dutton JJ. Coralline hydroxyapatite as an ocular implant. Ophthalmology 1991;98:370–377.
4. Oberfeld S, Levine MR. Diagnosis and treatment of complications of enucleation and orbital implant surgery. Adv Ophthalmic Plast Reconstr Surg 1990;8:107–117.
5. Baylis H, Shorr N. McCord C Jr, ed. Evisceration, enucleation, and exenteration. In: Oculoplastic Surgery. 1981:Raven Press, 313–326.
6. Kaltreider SA, Lucarelli MJ. A simple algorithm for selection of implant size for enucleation and evisceration: a prospective study. Ophthalmic Plast Reconstr Surg 2002;18:336–341.
7. Kaltreider SA, Jacobs JL, Hughes MO. Predicting the ideal implant size before enucleation. Ophthalmic Plast Reconstr Surg 1999;15:37–43.
8. Custer PL, Trinkaus KM. Volumetric determination of enucleation implant size. Am J Ophthalmol 1999;128:489–494.
9. Thaller VT. Enucleation volume measurement. Ophthalmic Plast Reconstr Surg 1997;13:18–20.
10. Karesh JW, Dresner SC. High-density porous polyethylene (Medpor) as a successful anophthalmic socket implant. Ophthalmology 1994;101:1688–1695; discussion 1695–6.
11. Ashworth JL, Rhatigan M, Brammar R, et al. A clinical study of the hydroxyapatite orbital implant. Eur J Ophthalmol 1997;7:1–8.
12. Verhoekx JSN, Rengifo Coolman A, Tse WHW, et al. A single- versus double-layered closure technique in anophthalmic surgery. Ophthalmic Plast Reconstr Surg 2017;33:329–333.
13. Tyers AG, Collin JR. Orbital implants and post enucleation socket syndrome. Trans Ophthalmol Soc U K 1982;102(pt 1):90–92.
14. Rubin PA, Popham J, Rumelt S, et al. Enhancement of the cosmetic and functional outcome of enucleation with the conical orbital implant. Ophthalmology 1998;105:919–925.
15. Shah CT, Hughes MO, Kirzhner M. Anophthalmic syndrome: a review of management. Ophthalmic Plast Reconstr Surg 2014;30:361–365.
16. Shams PN, Selva D; ANZSOPS Eyelid Retraction in Anophthalmia Survey Group. Upper eyelid retraction in the anophthalmic socket: review and survey of the Australian and New Zealand Society of Ophthalmic Plastic Surgeons (ANZSOPS). Ophthalmic Plast Reconstr Surg 2014;30:309–312.
17. Vistnes LM. Mechanism of upper lid ptosis in the anophthalmic orbit. Plast Reconstr Surg 1976;58:539–545.
18. Smit TJ, Koornneef L, Zonneveld FW, et al. Primary and secondary implants in the anophthalmic orbit. Preoperative and postoperative computed tomographic appearance. Ophthalmology 1991;98:106–110.
19. Soll DB. The anophthalmic socket. Ophthalmology 1982;89:407–423.
20. Sami D, Young S, Petersen R. Perspective on orbital enucleation implants. Surv Ophthalmol 2007;52:244–265.
21. Verhoekx JSN, Tse WHW, Rengifo Coolman A, et al. Complications following enucleations and subsequent oculoplastic surgeries. Ophthalmic Plast Reconstr Surg 2018;34:320–323.
22. Kaltreider SA. The ideal ocular prosthesis: analysis of prosthetic volume. Ophthalmic Plast Reconstr Surg 2000;16:388–392.
23. Custer PL, Trinkaus KM, Fornoff J. Comparative motility of hydroxyapatite and alloplastic enucleation implants. Ophthalmology 1999;106:513–516.
24. Custer PL, Maamari RN, Huecker JB, et al. Anophthalmic ptosis and the effects of enucleation on upper eyelid function. Ophthal Plast Reconstr Surg 2020. doi: 10.1097/IOP.0000000000001823. [Epub ahead of print]
25. Bekerman I, Gottlieb P, Vaiman M. Variations in eyeball diameters of the healthy adults. J Ophthalmol 2014;2014:503645.
26. Pearce E, Bridge H. Is orbital volume associated with eyeball and visual cortex volume in humans? Ann Hum Biol 2013;40:531–540.
27. Nunery WR, Heinz GW, Bonnin JM, et al. Exposure rate of hydroxyapatite spheres in the anophthalmic socket: histopathologic correlation and comparison with silicone sphere implants. Ophthalmic Plast Reconstr Surg 1993;9:96–104.
28. Kaltreider SA, Newman SA. Prevention and management of complications associated with the hydroxyapatite implant. Ophthalmic Plast Reconstr Surg 1996;12:18–31.
29. Tari AS, Malihi M, Kasaee A, et al. Enucleation with hydroxyapatite implantation versus evisceration plus scleral quadrisection and alloplastic implantation. Ophthalmic Plast Reconstr Surg 2009;25:130–133.
30. Moshfeghi DM, Moshfeghi AA, Finger PT. Enucleation. Surv Ophthalmol 2000;44:277–301.
© 2021 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc.