Abstract
Purpose :
Normal tissue complication probability (NTCP)-models could potentially be an important tool for understanding radiation-induced maculopathy after eye-preserving radiotherapy of malignant choroidal melanomas (CM). In this retrospective clinical study, we developed and compared NTCP-models for maculopathy for treatment with proton therapy (PT) and Ruthenium-106 brachytherapy (BT).
Methods :
We considered patients (pts) diagnosed with CM and primarily treated with PT (4154 pts, 52 Gy in 13 fractions) at Centre Antoine Lacassagne, Nice, or with BT (446 pts, 100 Gy to apex of tumor) at Rigshospitalet, Copenhagen.
Inclusion criteria were assessable data on maximum dose to macula and on macular outcome, reducing the cohorts to 3253 PT pts (treated 1991-2015) and 90 consecutive BT pts (treated 2005-2008). Presence of maculopathy was determined from visual acuity decrease or the presence of micro aneurysms, edema and/or ischemia in the macular area and was evaluated at each post-treatment visit. Eyeplan (for PT) and Plaque Simulator (for BT) were used as treatment planning systems and to extract macula dose data. Logistic regression was used to fit the NTCP-models using the statistical software R.
In a secondary analysis, we matched 109 PT pts (treated 2010-2011) to the BT pt cohort, using age, sex, tumor height, and distance from tumor to macula as match factors, and analyzed this selected PT cohort separately.
Results :
Median follow-up for BT, PT and selected PT cohorts were 7.4 years (interquartile range (IQR): 4.8-9.8), 4.3 years (2.0-8.0), and 3.1 years (2.2-3.7), respectively. Median maximum dose to macula for BT, PT and selected PT cohorts were 19 Gy (IQR: 5-53), 52 Gy (0-52), and 52 Gy (0-52), respectively. The prevalence of maculopathy was 4% for PT and 49% for BT. Maximum dose to macula and maculopathy were significantly correlated in the BT NTCP-model, with odds ratio (OR) for 10 Gy increase in dose 1.9 (95% CI: 1.3-2.8). A similar strong dose-response was not found for PT (OR: 1.1, 1.0-1.2). Neither did the comparable PT model show a significant effect (OR: 1.1, 0.8-1.6).
Conclusions :
Maculopathy rates were different for PT and BT. The NTCP-model for BT showed a dose-response relationship, in contrast to the PT-models. The difference might be due to varying follow-up times between the groups or that biological response might be different after treatments with the two modalities.
This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.