Efficacy and safety of secondary infero-medial orbital wall decompression after orbital fat decompression in Graves’ orbitopathy: A 12-year experience

Abstract

Purpose:

To evaluate the efficacy of secondary bony decompression following primary fat decompression.

Design:

Retrospective interventional case series.

Methods:

Setting: A tertiary university hospital.

Study Population:

A total of 27 eyes from 17 patients who underwent secondary inferomedial orbital wall decompression following primary orbital fat decompression were included. Patients with a history of trauma, extraocular muscle surgery, prior orbital irradiation, or eyelid or orbital inflammation within 6 months prior to decompression were excluded.

Intervention Procedure:

Secondary inferomedial orbital wall decompression performed for recurrent proptosis or compressive optic neuropathy after primary fat decompression.

Main Outcome Measure:

Changes in Hertel value, and best corrected visual acuity (BCVA).

Results:

Primary orbital fat decompression achieved an initial proptosis reduction of 3.4 ± 0.1 mm in 27 eyes. Secondary bony decompression was subsequently performed in 15 eyes from 9 patients with recurrent proptosis and in 12 eyes from 8 patients with dysthyroid optic neuropathy. Secondary bony decompression resulted in a reduction in Hertel measurement by 4.5 ± 2.1 mm, improving from 21.7 ± 2.8 mm to 17.3 ± 1.6 mm (P < 0.001) at a follow-up period of 40.6 ± 28.7 months. BCVA in patients with dysthyroid optic neuropathy improved from 0.43 ± 0.3 to 0.12 ± 0.15 logMAR (P = 0.005). Diplopia occurred in 12 of 17 patients (70%), including five with new-onset diplopia. Preoperative diplopia predicted postoperative diplopia in regression analysis.

Conclusion:

Secondary inferomedial orbital wall decompression following primary orbital fat decompression effectively reduces proptosis and improves vision.

Graves’ ophthalmopathy (GO) is a complex autoimmune dysfunction associated with inflammatory processes leading to orbital tissue expansion due to the overproduction of glycosaminoglycans and accumulation of orbital fat.[1] Rehabilitation surgery, involving orbital decompression to expand the volume of the orbit and thus treat complications such as proptosis and eyelid retraction, continues to be the primary treatment for inactive disease, particularly in moderate or severe cases that significantly impair quality of life.[2] In addition, surgical decompression is critical for preserving visual outcomes in compressive optic neuropathy in GO. Fat decompression surgery is an appealing option in orbital decompression due to its fewer complications compared to bony decompression.[3] Furthermore, the accurate prediction of postsurgical retrobulbar volume in fat decompression allows for a better correlation with the amount of proptosis correction.[4,5]

Although orbital fat decompression with preservation of the orbital wall can achieve predictable favorable results with minimal surgically induced adverse effects, there are patients who undergo a relapse or progression of their GO following initially successful fat decompression. The reactivation of GO is uncommon[6] and has been insufficiently documented particularly after orbital decompression. Recurring GO studies have determined a repeat bony decompression rate after prior bony decompression that ranges from 1.3% to 11%, depending on various surgical indications or definitions of the period of inactivity.[6,7,8,9] Of note, an American hospital-based study of 594 eyes that underwent orbital decompression surgery identified that bony decompression with adjunct fat removal is associated with decreased hazard (hazard ratio, HR = 0.32) of secondary decompression.[9]

Methods

The medical records of GO patients who underwent initial orbital fat decompression for proptosis and subsequently received secondary bony decompression at the National Taiwan University Hospital (Taipei, Taiwan) between 2009 and 2021 were retrospectively reviewed. The institutional Review Board approved this study (# 202307162RIND, 6.1.2023), and all procedures were performed in accordance with the principles stated in the Declaration of Helsinki.

All patients underwent initial orbital fat decompression via the transforniceal approach for proptosis. We excluded patients who had a history of trauma, extraocular muscle surgery, prior orbital irradiation, or eyelid and orbital inflammation within 6 months of decompression. Prolapsing extraconal and intraconal orbital fat were resected. Patient selection for primary fat decompression was based on clinical features, including prominent orbital fat prolapse and minimal extraocular muscle restriction. Orbital computed tomography (CT) was obtained in all patients as part of routine clinical evaluation, and CT findings were used to categorize patients into fat-predominant, muscle-predominant, and mixed type disease groups. The target volume of adipose tissue was quantified intraoperatively. Secondary orbital bony decompression was performed following recurrent proptosis, characterized by an increase of ≥2 mm in Hertel exophthalmometry measurements, or compressive optic neuropathy, which included pupillary abnormalities, decreased visual acuity, visual field defects, optic disc edema, and color vision impairment. The infero-medial orbital wall decompression was performed via the transforniceal approach for orbital bony decompression.

Data collected included age, gender, laterality of involvement, and smoking and thyroid status. Surgical outcomes of initial orbital fat decompression were evaluated. Postoperatively, changes in best corrected visual acuity (BCVA), Hertel exophthalmometry readings, time of follow-up, and development of new onset primary gaze diplopia lasting more than 3 months after secondary bony decompression were recorded. Specifically, changes in Hertel values were evaluated at baseline prior to the secondary bony decompression, 6 months postoperatively, and yearly thereafter. Diplopia noted within 30 degrees from primary gaze was also assessed at baseline and 6 months after secondary surgery; it was recorded as completely resolved, improved, unchanged, worsened, or newly induced. Post-bony decompression adjuvant ocular surgeries and complications were also documented.

Descriptive statistics for continuous variables were reported as mean ± SD and analyzed using SPSS software, V.24.0. Differences were analyzed by Student’s t-test. Correlation between variables was analyzed by Pearson’s correlation or the multiple logistic regression analysis. The BCVA was transformed into the logarithm of the minimal angle of resolution (log MAR) for the purpose of statistical analysis. A P value less than 0.05 was considered statistically significant.

Results

A total of 27 eyes in 17 patients (8 men and 9 women) with a mean age of 49.1 ± 12.0 years (range, 32–75 years) underwent secondary infero-medial orbital wall decompression following initial orbital fatty decompression. The indications for secondary orbital bony decompression were recurrent proptosis (n = 9 patients, 15 eyes, 55.6%) and dysthyroid optic neuropathy (n = 8 patients, 12 eyes, 44.4%). Six patients (35.3%) were current smokers.

For the primary orbital fat decompression, an average of 4.4 ± 1.3 mL of orbital fat was excised, resulting in an initial decrease in proptosis. However, during the follow-up period of 36.3 ± 52.6 months, proptosis remained at 21.7 ± 2.8 mm (P = 0.62) from the preoperative Hertel measurement of 21.1 ± 1.5 mm, reflecting a negligible change in Hertel value of 0.6 ± 1.3 mm. This finding reflects that although an initial postoperative reduction in proptosis was achieved after fat decompression, the Hertel value returned close to baseline during long-term follow-up in patients who subsequently required secondary decompression. In the secondary orbital bony decompression, on the other hand, the Hertel measurement reduced significantly by 4.5 ± 2.1 mm, from 21.7 ± 2.8 mm before surgery to 17.3 ± 1.6 mm (P < 0.001) after 40.6 ± 28.7 months postoperatively [Table 1]. For the 12 eyes of 8 patients with dysthyroid optic neuropathy, BCVA improved significantly from 0.43 ± 0.3 to 0.12 ± 0.15 logMAR (P = 0.005) after secondary bony decompression.

Table 1.

Clinical features of patients

		Overall		P
Sample size
Eyes		27
Patients		17
Age (years)
Mean±Standard Deviation		49.1±12.0
Range		32 - 75
Gender
Male		8
Female		9
Primary Orbital Fat Decompression
Follow-up (months)		36.3±52.6
Preoperative Hertel (mm)		21.1±1.5		0.62
Postoperative Hertel (mm)		21.7±2.8
Secondary Orbital Bony Decompression
Follow-up (months)		40.6±28.7
Preoperative Hertel (mm)		21.7±2.8		<0.001
Postoperative Hertel (mm)		17.3±1.6
Secondary Decompression Indication
Recurrent proptosis (eyes)		15
Optic neuropathy (eyes)		12
Best corrected visual acuity (logMAR)
Pre-bony decompression		0.24±0.31		0.04
Post-bony decompression		0.09±0.04
Current Smoker
Yes (n)		6
No (n)		11
Diplopia
Pre-bony decompression existed		7
Post-bony decompression		12
Improved		1
Same		1
Worsen		5
Newly Onset		5

log MAR=logarithm of the minimal angle of resolution; mm=milliliter; n=number of patients

Before secondary bony decompression, the 7 of 17 (41.2%) patients with existing diplopia had a mean Hertel value of 22.1 ± 3 mm, which was not statistically different from the mean value of 20.8 ± 1.7 mm (P = 0.1) among nondiplopic patients. Postoperatively, a total of 12 of 17 (70%) patients had diplopia, with one improving, one remaining unchanged, five worsening, and five developing new onset diplopia. The remaining five patients who did not subsequently have diplopia did not have it prior to surgery. Older patients at a mean age of 52.3 ± 11 years (P = 0.048) and existing diplopia prior to secondary bony decompression (P = 0.036) were predisposed to develop diplopia after secondary decompression. In contrast, postoperative Hertel values and BCVA were not significantly different for patients with and without diplopia (17.2 ± 1.7 mm vs 17.4 ± 1.2 mm, P = 0.5 and 0.096 ± 0.13 vs 0.085 ± 0.32 logMAR, respectively, P = 0.06). Furthermore, thyroid status, which was euthyroid at the time of secondary bony decompression in most patients (n = 14 of 17 patients, 82%), showed no significant difference (P = 0.3) between those who experienced diplopia (n = 9 of 12 patients) and those who did not (n = 5 of 5 patients). In addition, thyroid-stimulating hormone receptor antibody (TRAb) was detected in 8 of 17 (47%) patients, six of whom had postoperative diplopia. This was not statistically significant (P = 0.8) compared to the two nondiplopic patients [Table 2]. Post-bony decompression diplopia was significantly positively correlated with presurgical diplopia (P = 0.005) and age (P = 0.02), but not with gender, Hertel value, thyroid status, TRAb levels, or smoking status. A multivariable logistic regression model was implemented to evaluate the impact on postoperative diplopia, taking into account demographics, presurgical diplopia status, Hertel changes, smoking, and thyroid status. In the regression analysis, diplopia before bony decompression (B = 2.3, P < 0.001) predicted diplopia after surgery.

Table 2.

Clinical parameters in association with diplopia after secondary bony decompression

		Diplopia (+)		Diplopia (-)		P
Number		12		5
Age (years)		52.3		41.6		0.048
smoker (n)		5		1		0.6
Post bony decompression Hertel (mm)		17.2±1.7		17.4±1.2		0.5
Best corrected visual acuity (logMAR)		0.096±0.13		0.085±0.32		0.06
Existing diplopia prior to secondary surgery		7		0		0.036
Euthyroid		9		5		0.3
Thyroid-stimulating hormone receptor antibody		6		2		0.8

log MAR=logarithm of the minimal angle of resolution; mm=milliliter; n=number of patients

All 12 patients with diplopia underwent extraocular muscle surgery at least 6 months after secondary bony decompression, and 12 of 17 (70.6%) patients underwent eyelid retraction surgery. Based on the severity of upper and/or lower lid retraction, 12 eyes received upper eyelid surgery including anterior blepharotomy (n = 12) and mullerectomy (n = 2), whereas 8 eyes had lower eyelid surgery including correction of lower lid retraction (n = 8), hard palate (n = 2), and decellularized porcine-derived membrane (Tarsys implant) grafting (n = 2).

Discussion

To the authors’ knowledge, this is the first study to report surgical outcomes of secondary orbital bony decompression following primary orbital fat decompression. The reactivation of disfiguring proptosis or dysthyroid optic neuropathy in GO after orbital decompression is rarely reported. Our previous study is the first to show that the incidence of secondary bony decompression following primary solely fat decompression was 0.6% (10 of 1604 eyes),[3] which is much lower than the 7.6% (7 of 92 patients),[7] 9.1% (45 of 495 eyes),[8] and 13.8% (82 of 594 eyes)[9] reported in recent studies of primary bony decompression with[7,9] or without[8] fat removal. Although uncommon, reports of reactive GO have occurred after ocular surgeries such as cataract surgery, blepharoplasty, strabismus repair, or orbital decompression.[10] Of note, a hospital-based cross-sectional analysis of 415 cases of GO in an American study reported that surgeries were the most common identified triggers of recurrence, accounting for 12%, with cataract surgery at 5% and orbital decompression at 3%.[10] It is hypothesized that exposure to orbital soft tissues or trauma resulting from retrobulbar anesthesia or surgery may trigger the presence of self-antigens and perhaps orbital fibroblasts,[6] potentially leading to the reactivation of GO after these ocular surgeries. However, the majority of recurring GO, accounting for 77%, had no identifiable triggers.[10] In our practice, we elect to remove mostly the retro-equatorial intraconal orbital fat during the primary fat decompression. Intraconal fat, characterized by a high concentration of adipose-derived stem cells,[11] can differentiate into lipid-rich adipocytes or fibroblasts depending on the inflammation present in patients with recurrent GO.[12] We suggest that the removal of intraconal fat, and hence the important cellular differentiation source, may contribute to lower incidence of GO recurrence. This notion is further supported by a recent orbital bony and fat decompression study of 594 cases that showed adjunctive fat decompression along with bone removal decreased the hazard (HR: 0.32, P = 0.0004) for secondary decompression compared to solely primary bone extraction.[9] However, it is still unclear how fat removal or bony destruction in orbital decompression during the initial phase influences the frequency of the subsequent secondary occurrence.

Research has studied factors that increase the chance of recurrence, specifically smoking and thyroid status.[7,9,10] Compared to nonsmokers, smokers have been observed to have a higher rate of recurrence after primary orbital decompression.[7] In contrast, a Korean study that did not specify fat or bone removal found that smoking was nonsignificant for proptosis recurrence following rehabilitative orbital decompression.[13] There are, however, no data available on the impact of smoking on the recurrence of GO in patients following primary solely fat decompression. Our patients in this study had unchanged smoking behavior, and more than one third (35.3%) of patients were current smokers; however, we could not ascertain if smoking is an independent factor contributing to the reactivation of GO following primary fat decompression. At the time of the secondary decompression, most of our patients (n = 14 of 17 patients, 82%) were euthyroid. This result was consistent with a hospital-based Australian study[14] (n = 8 of 8 patients, 100%) of late GO and a Korean study[7] that reported more than half (n = 4 of 7 patients, 57%) of recurrences after primary decompression was euthyroid. Given that the majority of orbital decompressions are performed on euthyroid individuals, reactivation is typically unpredictable. In our study, TRAb was detected in about half of the patients (n = 8 of 17, 47%). However, quantitative measurements were not collected; hence, we cannot determine if any specific preoperative TRAb acts as a prognostic marker for recurrence following orbital decompression, as shown in a Korean study.[13]

Our current study showed that out of 17 patients who required repeat decompression after primary fat decompression, 41.2% had diplopia prior to the secondary procedure, and an additional 5 developed diplopia after the repeat, resulting in a total of 12 out of 17 (70%) patients having diplopia postoperatively. These results were different from our previous study utilizing the same surgical technique, which revealed a much lower occurrence of newly onset diplopia after fat decompression surgery (n = 23 of 692 patients, 3.3%,[3] and n = 3 out of 105 patients, 2.8%[5]). The incidence of pre- and postoperative diplopia following repeat orbital decompression after primary fat removal was high in our study. Another American study demonstrated results that aligned with ours, revealing 40% of patients experienced diplopia prior to the repeat decompression following primary bone removal, and another 24% experienced newly developed diplopia following the repeat decompression.[8] Although numerous hypotheses have been proposed to explain the induction of diplopia following orbital decompression, regardless of the removal of fat or bone, none have been well established. A study showed that preoperative Gorman diplopia score associated with predominant extraocular muscle enlargement was a significant factor for reactivation of GO after orbital decompression.[7] The retention of an antigenic component from muscle tissue with autoreactive lymphocytes increases inflammatory cytokines following either fat or bony decompression, potentially triggering the reactivation of the immune response.[15] This might help explain why pre- and postoperative diplopia was so common in our and other studies’ secondary decompression patients. Our findings revealed that pre-existing diplopia prior to secondary bony decompression significantly predicted the occurrence of diplopia after repeat decompression, particularly in the absence of muscle intervention, thereby supporting this theory. Furthermore, we believe that the pre-existing diplopia, which indicates potentially more fibrotic changes resulting from prior decompression surgery, may enhance the probability of new-onset diplopia after secondary orbital decompression.

One limitation of this study is the lack of strict selection criteria to differentiate GO patients with either muscle- or fat-predominant disease. Orbital CT was obtained in all patients as part of routine clinical evaluation, and CT findings were used to categorize patients into fat-predominant, muscle-predominant, and mixed type disease groups. As a result, some patients with mixed fat and muscle disease may have undergone primary fat decompression. The efficacy of primary fat decompression in reducing proptosis has been demonstrated in case series with follow-ups ranging from 3 to 5.9 mm at 1–3 years.[16,17] Our previous study demonstrated consistent long-term efficacy of orbital fat decompression in reducing proptosis of 4.6 ± 1.3 and 4.2 ± 1.4 mm at 3–5 and 5–10 years, respectively.[3] In contrast, this study disclosed minimal improvement of proptosis despite an adequate removal of orbital fat and an initial reduction in proptosis. We suspect that the lack of strict selection only for fat-predominant GO patients contributed to this limited outcome, with in proptosis regression occurring after initial fat decompression. Orbital CT scans were obtained for all patients, and imaging findings were used to determine whether muscle enlargement, fat predominance, or mixed type was present in patients with disfiguring exophthalmos. The subsequent development of dysthyroid optic neuropathy in 8 of 17 patients suggests that hypertrophied extraocular muscles at the orbital apex may have been a contributing factor.

No consensus has been established about the most effective surgical technique for primary or secondary decompression. While some surgeons prefer a balanced approach involving medial and lateral wall decompression as their primary strategy,[18] our recent 30 year study revealed that inferomedial orbital wall decompression demonstrated an excellent success rate with satisfactory proptosis reduction and improved visual outcome in patients with disfiguring proptosis or dysthyroid optic neuropathy.[19] Pre- and postoperative comparisons of BCVA in our study, as well as in other secondary decompression studies, showed statistically significant improvement for patients with compressive optic neuropathy.[20] However, there is limited research on the effects of Hertel correction across different secondary surgical decompression techniques.

Our study showed that secondary inferomedial orbital wall decompression led to a Hertel reduction of 4.5 ± 2.1 mm after 40.6 ± 28.7 months of follow-up. This may suggest that the procedure was more effective at reducing proptosis than mixed surgical approach techniques for secondary decompression, which typically show a 1.6 mm improvement.[8] We postulate that, although there are fibrotic changes following primary decompression surgeries, secondary inferomedial orbital wall decompression creates a large sinus space for the long-term redistribution of residual fat tissue and leads to a greater decrease in proptosis compared to other surgical approaches.[19] However, due to the varying severity and clinical nature of recurrent GO, further research is necessary to validate these observations. Although secondary orbital bony decompression is effective in reducing proptosis and improving vision in patients with recurrent proptosis or compressive optic neuropathy, it may still be associated with an uncertain risk. To prevent recurrence, further prospective comparison studies should aim to focus on fat predominance cases to examine independent parameters linked to the reactivation of GO following initial fat decompression.

Conclusion

To our knowledge, there is limited literature regarding secondary decompression following a primary solely orbital fat decompression. In this study, we describe the surgical outcomes of patients who initially underwent orbital fat decompression and subsequently received secondary bony decompression after experiencing recurrent proptosis or compressive optic neuropathy.

Data availability statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Author contributions

A.C.: research execution, data analysis and/or interpretation, manuscript preparation and review and editing; Y.T.F.: research execution, data analysis and/or interpretation, manuscript preparation; Y.H.W.: manuscript preparation; A.Y.C.: manuscript preparation; H.B.: manuscript preparation; E.S.Y.: manuscript preparation; S.L.L.: research design, research execution, data analysis and/or interpretation, supervision, manuscript preparation and review and editing.

Conflicts of interest

There are no conflicts of interest.

Funding Statement

Nil.