Primary transpupillary thermotherapy for choroidal melanoma in 391 cases: importance of risk factors in tumor control

Purpose: To report the long-term outcome of primary transpupillary thermotherapy (TTT) for choroidal melanoma.

Design: Retrospective review of medical records.

Participants: We included 391 patients with choroidal melanoma treated between 1995 and 2012 at the Oncology Service, Wills Eye Hospital, Philadelphia.

Methods: We delivered TTT with an infrared diode laser.

Main outcome measures: Local tumor recurrence, Snellen visual acuity after TTT, and distant metastasis.

Results: Of 391 patients, 311 (80%) were treated from 1995 to 2000 and 80 (20%) from 2001 to 2012. Tumors in the 2001 to 2012 group were ultrasonographically thinner (2.2 vs. 2.7 mm), more distant from the optic disc (3.2 vs. 2.5 mm) and foveola (4.0 vs. 2.0 mm), were less often located in the macular area (14% vs. 40%), and had lower rates of acoustic hollowness on B-scan ultrasonography (63% vs. 84%), subretinal fluid (58% vs. 90%), and orange pigment (50% vs. 70%). Kaplan-Meier estimates for tumor recurrence in the 1995 to 2000 group were 29% at 5 years and 42% at 10 years, whereas estimates for tumor recurrence in the 2001-2012 group were 11% at 5 years and 15% at 10 years. Of 108 recurrent tumors 20 were controlled with additional TTT and 62 required plaque radiation (n=60) or proton beam radiation (n=2), with enucleation necessary in 26 patients. Tumor recurrence correlated with the number of high-risk tumor features: 10-year recurrence was 18% in those with 1 or 2 risk factors, 35% in those with 3 to 5 factors, and 55% in those with 6 or 7 factors. On multivariate analysis, features predictive of tumor recurrence were presence of symptoms (P<0.001), shorter distance between the tumor and the optic disc (P=0.026), subretinal fluid (P=0.035), thickness of residual tumor scar (P<0.001), and elevation of residual tumor scar (P<0.001). The only factor predictive of extraocular tumor extension was intraocular tumor recurrence after TTT treated with additional TTT (P=0.007). Presence of orange pigment before TTT (P=0.019), tumor recurrence (P=0.002), and extraocular tumor extension (P=0.017) were predictive of distant metastasis.

Conclusion: This study shows a direct correlation between a larger number of high-risk tumor features and higher rates of tumor recurrence after primary TTT of (small) choroidal melanoma. We advise that, when possible, small choroidal melanomas with multiple risk factors be treated with methods other than TTT.