Clinical impact of MRI assisted dose volume adaptation and dose escalation in brachytherapy of locally advanced cervix cancer.

BACKGROUND: To investigate the clinical impact of MRI based cervix cancer brachytherapy combined with external beam radiochemotherapy applying dose volume adaptation and dose escalation in a consecutive group of patients with locally advanced cervix cancer.
METHODS: In the period 1998-2003, 145 patients with cervix cancer stages IB-IVA were treated with definitive radiotherapy +/- cisplatin chemotherapy. Median age was 60 years. In 67 patients, the tumour size was 2-5 cm, in 78 patients it was >5 cm. In 29 cases the standard intracavitary technique was combined with interstitial brachytherapy. Total prescribed dose was 80-85 Gy (total biologically equivalent dose in 2 Gy fractions). Since 2001, MRI based treatment planning integrated systematic concepts for High Risk Clinical Target Volume (HR CTV) and organs at risk (OAR), biological modelling, Dose-Volume-Histogram analysis, dose-volume-adaptation (D90, D 2 cm(3)), and dose escalation, if appropriate and feasible.
FINDINGS: Dose volume adaptation was performed in 130/145 patients. The mean D90 during the whole period was 86 Gy, with a mean D90 of 81 Gy and 90 Gy during the first and second period, respectively (p<<0.01). Median follow-up was 51 months. Complete remission at 3 months was achieved in 138/145 patients (95%). Actuarial continuous complete remission for true pelvis (CCRtp) was 88% at 3 years. For tumours 2-5 cm, CCRtp was 96% both in 1998-2000 and 2001-2003. For tumours >5 cm it was 71% in 1998-2000 and 90% in 2001-2003 (p=0.05). Progression free survival (PFS) for true pelvis (local control) was 85%, PFS for distant metastases was 80%, both at 3 years. Local control for tumours >5 cm was 64% in 1998-2000 and 82% in 2001-2003 (p=0.09) and 100% and 96%, respectively, for tumours 2-5 cm. PFS for distant metastases remained the same during the two treatment periods with 79% and 80%. Overall survival (OS) was 58%, and cancer-specific survival (CSS) was 68% at 3 years. In the two different periods improvement in OS was from 53% to 64% (p=0.03) and in CSS from 62% to 74% (p=0.13). Improvement occurred only in tumours >5 cm: OS 28% versus 58% (p=0.003); CSS 40% versus 62% (p=0.07). Actuarial late morbidity rate (LENT SOMA, grades 3 and 4) at 3 years was gastrointestinal 4%, urinary 4% and vaginal 5% (stage IIA/IIIA). Gastrointestinal and urinary late morbidity (G3,G4) was 10% in 1998-2000 and 2% in 2001-2003.
INTERPRETATION: In locally advanced extensive cervix cancer, local control of > or = 85% can be achieved with low treatment related morbidity (G3/G4), when exploiting the potential of MRI based 3D treatment planning including dose volume adaptation and dose escalation and a combined intracavitary/interstitial brachytherapy, if appropriate. A significant impact of this improvement of local control on survival is to be expected. For locally advanced limited disease the MRI based approach will likely result in assuring excellent local control (> or = 95%) and in minimizing treatment related morbidity.