MRI-guided 3D optimization significantly improves DVH parameters of pulsed-dose-rate brachytherapy in locally advanced cervical cancer

Purpose: To compare dose-volume histogram parameters of standard Point A and magnetic resonance imaging-based three-dimensional optimized dose plans in 21 consecutive patients who underwent pulsed-dose-rate brachytherapy (PDR-BT) for locally advanced cervical cancer. Methods and Materials: All patients received external beam radiotherapy (elective target dose, 45 Gy in 25-30 fractions; tumor target dose, 50-60 Gy in 25-30 fractions). PDR-BT was applied with a tandem-ring applicator. Additional ring-guided titanium needles were used in 4 patients and a multichannel vaginal cylinder in 2 patients. Dose planning was done using 1.5 Tesla T-1-weighted and T-2-weighted paratransversal magnetic resonance imaging scans. T-1-weighted visible oil-containing tubes were used for applicator reconstruction. The prescribed standard dose for PDR-BT was 10 Gy (1 Gy/pulse, 1 pulse/h) for two to three fractions to reach a physical dose of 80 Gy to Point A. The total dose (external beam radiotherapy plus brachytherapy) was normalized to an equivalent dose in 2-Gy fractions using alpha/beta = 10 Gy for tumor, alpha/beta = 3 Gy for normal tissue, and a repair half-time of 1.5 h. The goal of optimization was dose received by 90% of the target volume (D-90) of >= 85 Gy(alpha/beta 10) in the high-risk clinical target volume (cervix and remaining tumor at brachytherapy), but keeping the minimal dose to 2 cm 3 of the bladder and rectum/sigmoid at < 90 and < 75 Gy(alpha/beta 3), respectively. Results: Using three-dimensional optimization, all dose-volume histogram constraints were met in 16 of 21 patients compared with 3 of 21 patients with two-dimensional library plans (p < 0.001). Optimization increased the minimal target dose (D-100) of the high-risk clinical target volume (p < 0.007) and decreased the minimal dose to 2 cm(3) for the sigmoid significantly (p = 0.03). For the high-risk clinical target volume, D-90 was 91 +/- 8 Gy(alpha/beta 10) and D-100 was 76 +/- 5 Gy(alpha/beta 10)). The minimal dose to 2 cm(3) for the bladder, rectum, and sigmoid was 73 +/- 6, 67 +/- 6, and 69 +/- 6 Gy(alpha/beta 3), respectively. Conclusion: The results of our study have shown that magnetic resonance imaging-guided optimization of PDR-BT for locally advanced cervical cancer significantly improved the dose-volume histogram parameters. (c) 2008 Elsevier Inc.