Factors Predictive of Tumor Growth, Tumor Decalcification, Choroidal Neovascularization, and Visual Outcome in 74 Eyes With Choroidal Osteoma

Objective  To evaluate choroidal osteoma for tumor growth, tumor decalcification, related choroidal neovascularization, visual acuity loss, and poor visual acuity.

Design  Retrospective nonrandomized single-center case series.

Setting  Ocular Oncology Service at Wills Eye Hospital, Thomas Jefferson University, Philadelphia, Pa.

Participants  There were 74 eyes of 61 patients with choroidal osteoma evaluated between January 1, 1977, and January 1, 2003.

Main Outcome Measures  The 5 outcome measures included tumor growth, tumor decalcification, related choroidal neovascularization, visual acuity loss of 3 or more Snellen lines, and poor visual acuity of 20/200 or worse.

Results  At 5 and 10 years, Kaplan-Meier analysis revealed tumor growth in 22% and 51% of eyes, tumor decalcification in 28% and 46% of eyes, choroidal neovascularization in 31% and 31% of eyes, visual acuity loss in 26% and 45% of eyes, and poor visual acuity in 45% and 56% of eyes, respectively. The clinical factor predictive of tumor growth was absent overlying retinal pigment epithelial alterations. The factor predictive of decalcification was irregular tumor surface. Of the 15 tumors that showed partial decalcification at the first visit, there was no further tumor growth in any case. Of the remaining 12 tumors that later developed decalcification, tumor growth, if it occurred, was along the margin opposite the decalcification. No tumor showed growth in the region of decalcification. Factors predictive of choroidal neovascularization included irregular tumor surface and subretinal hemorrhage. In 6 eyes that had both choroidal neovascularization and tumor decalcification, the neovascularization was detected prior to or at the same time as the decalcification. The factor predictive of visual acuity loss was presence of subretinal fluid whereas the factors predictive of poor visual acuity included symptoms and tumor decalcification. A comparison of eyes with subfoveal vs extrafoveal choroidal osteoma showed poor visual acuity in 15 (34%) of 44 eyes and 3 (10%) of 30 eyes, respectively. Eyes with decalcified choroidal osteomas manifested poor visual acuity in 13 (48%) of 27 eyes whereas those with nondecalcified tumors showed poor visual acuity in 5 (11%) of 47 eyes.

Conclusions  Choroidal osteoma showed evidence of growth in 51% of eyes and decalcification in nearly 50% of eyes by 10 years. Tumors with any degree of decalcification at the initial visit showed no further growth. Overall, poor visual acuity of 20/200 or worse was found in 56% of eyes by 10 years, and decalcified subfoveal choroidal osteomas displayed a particularly poor visual prognosis.

Choroidal osteoma is a benign intraocular tumor composed of mature bone that typically replaces the full thickness of the choroid. This tumor classically manifests as an orange-yellow plaque deep to the retina in the juxtapapillary or macular region.^1-4 It most often occurs as a unilateral condition in teenaged or young adult females. Unfortunately, the etiology and pathogenesis of this tumor is poorly understood.

Since the initial descriptions of choroidal osteoma in 1978,^1-3 most studies have focused on the clinical and diagnostic features of this tumor as well as methods of management of related choroidal neovascularization. It was later recognized that choroidal osteoma showed evolution with change in clinical features over many years, and cases of tumor growth,^5-9 tumor decalcification or involution,^10-13 and related choroidal neovascularization^14-19 were documented. In 1998, Aylward et al²⁰ provided observations on long-term outcome in a series of 36 patients with choroidal osteoma. They indicated that there was a moderate risk for development of vision-threatening choroidal neovascularization in eyes with choroidal osteoma.

In this study, we reviewed a series of 61 patients with choroidal osteoma to delineate the likelihood of tumor growth, tumor decalcification, choroidal neovascularization, visual acuity loss, and poor visual acuity, and to identify clinical factors predictive of each of these outcomes as well as the relationship of decalcification to these outcomes.

All of the patients with choroidal osteoma who were examined and managed in the Ocular Oncology Service, Wills Eye Hospital, Thomas Jefferson University, Philadelphia, Pa, between January 1, 1977, and January 1, 2003, were included in this study. Choroidal osteoma was defined as a yellow-orange mass deep to the retinal pigment epithelium (RPE) and occupying the choroid with well-defined margins and bone density on ultrasonography or computed tomography.^21,22 Cases of sclerochoroidal calcification,^23,24 a condition often confused with choroidal osteoma, were excluded from this survey. The clinical details of each patient and his or her tumor were collected retrospectively by record review.

The reviewed patient data included age, race, sex, systemic illnesses, systemic medications, previous ocular diagnosis and treatment, family history of ocular problems, and referring diagnosis. The ocular data at initial examination included symptoms, tumor laterality, visual acuity, intraocular pressure, iris color, and anterior segment findings. The tumor data at initial examination included tumor location (juxtapapillary, macula), quadrant location (superior, temporal, inferior, nasal, macula), proximity to optic nerve and foveola (millimeters measured by indirect ophthalmoscopy), tumor diameter (millimeters measured by indirect ophthalmoscopy), tumor thickness (millimeters measured by ultrasonography), tumor color, basal configuration, intrinsic vascularity, and surface appearance (irregular, smooth). Regions of tumor decalcification or involution were noted. Tumor decalcification or involution was defined as areas within the osteoma displaying loss of bone, RPE alterations, and visibility of underlying choroidal vessels. Data regarding associated findings included RPE alterations, subretinal fluid, subretinal hemorrhage, subretinal fibrosis, and choroidal neovascularization. Visual acuity at last examination was recorded.

The 5 outcomes included tumor growth, tumor decalcification, choroidal neovascularization, visual acuity loss of 3 Snellen lines or more, and poor final visual acuity. The time to onset of each of these outcomes was recorded.

The clinical data were then analyzed with regard to the 5 outcomes of tumor growth, tumor decalcification, choroidal neovascularization, visual acuity loss of 3 or more Snellen lines, and poor visual acuity of 20/200 or worse. The effect of each individual clinical variable recorded at the time of the first visit of the patient at the Ocular Oncology Service on each of the outcomes was analyzed by a series of univariate binary logistic regressions. The correlation among the variables was determined using Pearson correlations. All of the variables were analyzed as discrete variables (continuous variables were analyzed by grouping them into discrete categories). Variables that were significant on a univariate level (P≤.05) were entered first into the multivariate logistic regression analysis. For variables that showed a high degree of correlation, only 1 variable from the set of associated variables was entered at a time into subsequent multivariate models. A final multivariate model tested variables that were identified as significant predictors (P≤.05, forward stepwise [conditional or 95% confidence interval of the relative risk]) from the initial model as well as variables deemed clinically important for each of the outcomes. In the final model, a predictor was considered a significant risk factor if the 95% confidence interval of its relative risk did not contain a risk value of 1. Kaplan-Meier life-table analysis was used to analyze the time from the initial examination at the Ocular Oncology Service to each of the 5 outcomes.

The demographic information of the 61 patients is listed in Table 1. The median age at the first visit was 25 years, 41 patients (67%) were female, and 13 patients (21%) had bilateral tumors. The referring diagnoses are listed in Table 2. Approximately 50% of the patients were referred with the correct diagnosis whereas the other 50% had mistaken diagnoses or no referral diagnoses.

The clinical features at the initial visit are listed in Table 3. There were 21 patients (34%) without symptoms. Macular involvement with tumor was seen in 42 eyes (57%). Poor visual acuity of 20/200 or worse was found in 15 eyes (20%). Evidence of choroidal neovascularization was noted at the first examination in 14 eyes (19%).

A summary of the 5 outcomes is listed in Table 4. By 10 years, 51% of tumors showed evidence of growth, 46% showed decalcification, and 31% showed related choroidal neovascularization (Figures 1, 2, 3, and 4). By 10 years, 45% of eyes had lost 3 or more lines of visual acuity, and 56% had poor visual acuity of 20/200 or worse (Figure 5). The location and rate of tumor growth is shown in Table 5. In the 13 tumors that grew, the mean growth rate was 0.37 mm/y. The appearance and progression of tumor decalcification is shown in Table 6. Decalcification was detected initially in the central region in 10 eyes (37%) and in the peripheral region in 13 eyes (48%). During follow-up, continued decalcification was seen in 9 eyes (33%), and it occurred along the peripheral rim in 6 eyes (67%) and in the central region in 3 eyes (33%). Of those with progressive decalcification, approximately 22% of the tumor was decalcified by 5 years whereas 58% of the tumor was decalcified at 20 years. Of the 15 tumors that showed partial decalcification at the first visit, there was no further tumor growth in any case. Of those 12 that developed decalcification during follow-up, regions of tumor growth were only detected on the side of the tumor opposite the decalcification. The location and time of onset of choroidal neovascularization are listed in Table 7. In all of the cases, the choroidal neovascularization was diagnosed at the same time or prior to the decalcification. In the 4 eyes that showed choroidal neovascularization in the bed of decalcification (Table 7), previous laser photocoagulation had not been performed.

Predictive factors at the initial visit are listed for tumor growth in Table 8, tumor decalcification in Table 9, related choroidal neovascularization in Table 10, visual acuity loss in Table 11, and poor visual acuity in Table 12. A summary of final visual acuity results in patients with or without subfoveal involvement of tumor and with or without tumor decalcification is displayed in Table 13. Eyes with decalcified choroidal osteomas manifested poor visual acuity in 13 (48%) of 27 eyes whereas those with nondecalcified tumors showed poor visual acuity in 5 (11%) of 47 eyes.

Choroidal osteoma is a relatively rare intraocular tumor. A PubMed computer search for the words choroid osteoma yielded approximately 100 articles, most of which were case reports or small series on clinical features, diagnostic imaging, or therapy. Most authors have had experience with only 1 or a few cases of this condition during their career. The first relatively large series of patients with choroidal osteoma was reported by Gass⁵ in 1979. In that study of 15 affected patients, he provided new observations that choroidal osteoma could occur in males, could be sequential rather than simultaneous, and could show marked enlargement. A later assessment of his enlarged series of 36 patients with longer follow-up was described by Aylward et al.²⁰ The rarity of choroidal osteoma is further emphasized in our current series, as only 61 patients with choroidal osteoma were identified in our ocular oncology practice over a 26-year period.

Choroidal osteoma is generally diagnosed in young symptomatic patients.^4,25,26 In our series, the mean age at diagnosis was 26 years. However, patients as young as 2 months and as old as 67 years were found. Intraocular calcification in an otherwise normal eye of an older adult should raise suspicion for idiopathic sclerochoroidal calcification, a form of benign asymptomatic dystrophic calcification, rather than choroidal osteoma.^23,24 Sclerochoroidal calcification tends to occur bilaterally in older patients, and it involves the sclera and outer choroid along the superotemporal or inferotemporal retinal vascular arcades in a multifocal ringlike pattern. Unlike choroidal osteoma, the visual acuity is typically not affected by sclerochoroidal calcification.

Choroidal osteoma can lead to important vision-threatening problems. Tumor growth, tumor decalcification, and related choroidal neovascularization can contribute to substantial visual loss and poor visual acuity. In our analysis, the 10-year probability was 51% for tumor growth, 46% for tumor decalcification, 31% for related choroidal neovascularization, 45% for visual acuity loss, and 56% for poor visual acuity. Since this condition typically occurs in otherwise healthy, young patients, most can anticipate experiencing 1 or many of these outcomes. Similar to our results, Aylward et al²⁰ found that the 10-year probability was 41% for tumor growth, 47% for related choroidal neovascularization, and 58% for poor visual acuity. They did not assess for tumor decalcification or visual acuity loss of 3 lines or more.

Growth of choroidal osteoma is usually slow, with evolution photographically documented over many months or years.^7,20 In our analysis, those eyes with tumor growth showed progression in mean basal diameter of 0.37 mm/y. Tumor growth appeared to be random along the margins, with no predilection for any specific margin. The only exception was that growth did not occur in any case on the margin of the tumor that showed decalcification. The absence of RPE changes overlying the tumor was the only factor in our series that was predictive of tumor growth. Lack of RPE changes implies a less chronic condition, perhaps early in evolution. Mizota et al⁸ described rapid evolution of choroidal osteoma in a 3-year-old girl, with quadrupling of diameter over 3 years. In that case, there was no photographic evidence of RPE alterations over the growing tumor.

Decalcification of choroidal osteoma was first described by Trimble et al¹⁰ in 1988. They reported a case in which the tumor grew over 5 years, then began to decalcify, and later developed choroidal neovascularization in the sixth year. By the ninth year, the tumor had completely disappeared, leaving a bed of RPE and choriocapillaris atrophy and 20/100 visual acuity. In our series of 74 eyes with choroidal osteoma, 15 lesions showed partial decalcification at the initial visit, none of which showed growth on follow-up. It is unknown how long these tumors were present or decalcified, but this implies that decalcification signifies not only involution of the calcified portion but also stabilization of the tumor scar at that site.

Others have observed spontaneous decalcification¹² or laser-induced decalcification,^11,13,16 often with poor long-term vision. In our series, decalcification of choroidal osteoma led to gradual atrophy of the tumor, but unfortunately, it was commonly found with poor vision. Patients with decalcified tumors showed poor visual acuity in 13 (48%) of 27 eyes as compared with only 5 (11%) of 47 eyes of those without decalcification. Subfoveal choroidal osteoma led to poor visual acuity in 10 (48%) of 21 eyes if the tumor was decalcified and in only 5 (22%) of 23 eyes when the tumor was not decalcified. Importantly, decalcification was a significant factor in the analysis of risks for poor long-term visual acuity. The relationship of decalcification to poor visual acuity is unknown. Decalcification commonly occurs concurrently with overlying RPE alterations and choriocapillaris atrophy, both of which could lead to photoreceptor degeneration and poor visual acuity. As shown in Figure 2 and Figure 5, optical coherence tomography confirmed loss of photoreceptors overlying subfoveal choroidal osteoma with poor visual acuity, but with preservation of adjacent photoreceptors.

Choroidal neovascularization was associated with choroidal osteoma by the 1-year follow-up in 21% of eyes, by the 5-year follow-up in 31% of eyes, by the 10-year follow-up in 31% of eyes, and by the 20-year follow-up in 46% of eyes. Tumors with overlying hemorrhage and irregular surface were at the greatest risk for development of choroidal neovascularization. Disruption of the RPE and thinning or loss of the Bruch membrane and choriocapillaris might contribute to the development of choroidal neovascularization. Choroidal neovascularization occurred before or at the same time as detection of decalcification in all of the 6 cases in which both were present. Choroidal neovascularization and decalcification may be related. In fact, choroidal neovascularization occurred in the bed of decalcification in 4 (67%) of 6 eyes. Previous articles^14-19 have documented various methods of treating choroidal neovascularization, including laser photocoagulation, photodynamic therapy, and removal of the membrane via subretinal extraction. Despite therapy, choroidal neovascularization often leads to poor vision.

Visual acuity loss of 3 lines or more occurred in 45% of patients by 10 years and in 59% of patients by 20 years (Table 4). Poor visual acuity of 20/200 or worse was found in 56% of patients by 10 years and in 62% of patients by 20 years. Many factors contribute to final visual acuity, such as tumor location, choroidal neovascularization, subretinal fluid, and overlying RPE alterations. In this analysis, it was noted that decalcification of the tumor was a risk factor for poor long-term visual acuity. Those eyes with subfoveal decalcified tumor showed poor visual acuity in 10 eyes (48%), usually with visible RPE and choriocapillaris atrophy. Those with extrafoveal decalcified tumors showed poor visual acuity in 3 eyes (50%), all of which had circumpapillary tumors with resultant optic atrophy.²⁷

In summary, choroidal osteoma is a benign intraocular tumor with classic clinical features. Despite the benign appearance, choroidal osteoma can lead to tumor enlargement, tumor decalcification, and related choroidal neovascularization and can cause profound visual acuity loss over many years. Future therapies should be directed toward stabilization of the tumor, prevention of decalcification of subfoveal tumors, or stimulating tumor decalcification at an early point, particularly in cases where the fovea is threatened by an advancing margin of the osteoma.

Correspondence: Carol L. Shields, MD, Ocular Oncology Service, Suite 1440, Wills Eye Hospital, 840 Walnut St, Philadelphia, PA 19107 (carol.shields@shieldsoncology.com).

Submitted for Publication: July 20, 2004; final revision received January 24, 2005; accepted January 25, 2005.

Financial Disclosure: None.

Funding/Support: This work was supported by the Macula Foundation, New York, NY (Dr C. L. Shields), Mellon Charitable Giving from the Martha W. Rogers Charitable Trust, Mellon Financial Corp, Pittsburgh, Pa (Dr C. L. Shields), the Rosenthal Award of the Macula Society, Cleveland, Ohio (Dr C. L. Shields), the Eye Tumor Research Foundation, Philadelphia, Pa (Dr C. L. Shields), a donation from Michael Ratner, Bruce Ratner, and Ellen Ratner, New York (Drs C. L. Shields and J. A. Shields), the Chinese Education Association for International Exchanges, Shandong Branch, Jinan, People’s Republic of China (Dr Sun), and the Paul Kayser International Award of Merit in Retina Research, Retina Research Foundation, Houston, Tex (Dr J. A. Shields).