Evaluation Of Dose-Volume Metrics Of An 18f-Fdg Positron Emission Tomography Adaptive Treatment Planning Protocol For Gynecological Malignancies

The purpose of this study was to assess dose-volume metrics of a PET adaptive treatment planning protocol. Our hypothesis was that adaptive re-planning using the intra-treatment PET-CT would produce significantly lower dose-volume metrics for the normal tissue structures, while maintaining target coverage. 50 women with PET visible malignancies were enrolled in a prospective, IRB-approved 18F-FDG PET adaptive treatment planning protocol. In this study, 21 cervical patients were assessed. A pre-treatment PET-CT was acquired for treatment planning. Prescription dose to the primary PTV was 45 Gy, with a 55-70 Gy boost to the nodes. Primary and boost plans were planned sequentially or as a simultaneous integrated boost. An intra-treatment PET-CT was acquired between 30-36 Gy. The pre- and intra-treatment CTs were registered using rigid registration based on bony landmarks. Normal tissue and target were re-contoured and re-planning was performed to generate the adapted plan. To assess the impact of the adapted plan on dose-volume metrics, an unadapted plan was generated by applying the original plan to the intra-treatment CT for the remaining fractions. Clinically meaningful metrics were evaluated including: volumes; D2cc of the bladder, bowel, rectum, kidneys, femoral heads, and sigmoid; D10cc of the bladder and rectum; D195cc of the bowel; D0.03cc of the spinal cord; and D90%, D95%, and D98% of the primary GTV and nodes and of the primary and boost PTVs. Statistical significance between the adapted and unadapted plans was assessed using the Wilcoxon signed-rank test (two-tail, P<0.05). Median relative volume differences were significantly lower for the GTV (primary: -48 ± 9.2%, nodes: -57.9 ± 10.3%, P<0.001) and the PTV (primary: -0.9 ± 0.8%, P = 0.04; boosts: -27.8 ± 16.1%, P = 0.02), while volumes for most normal tissue structures were similar when comparing the pre- to intra-treatment CT contours. Adapted plans generally resulted in lower dose-volume metrics to normal tissue, while maintaining target coverage for the remaining fractions. Median D2cc was significantly different for bowel (13.9 vs 17.9 Gy; P<0.001), bladder (12.9 vs 14.0 Gy; P<0.001), rectum (13.8 vs 14.8 Gy; P = 0.002), femoral heads (9.4 vs 9.7 Gy; P<0.03), kidneys (5.1 vs 5.9 Gy; P<0.04), and sigmoid (13.9 vs 14.1 Gy; P = 0.002). D10 cc for bladder and rectum were significantly lower, as was D195cc for the bowel. D0.03cc for the cord remained the same. Median primary and boost PTV D95% and D98% were similar for both plans. Adaptive treatment planning protocols for gynecological malignancies present an opportunity to ensure that dose to the normal tissue is kept within the limits or even reduced while maintaining the prescription dose coverage.