Image Guidance Affects Biochemical Outcome For Postprostatectomy Radiation Therapy

The use of daily image guidance (IGRT) for prostate radiation therapy now pervades clinical practice. While evidence exists for definitive prostate radiation therapy that IGRT reduces acute toxicity while maintaining biochemical control, no comparison of toxicity or biochemical outcome for IGRT versus non-daily imaging exists in the post-operative setting. In this retrospective analysis, we examine the effect of IGRT on acute and late toxicity and biochemical control for men treated at a single institution with post-operative prostate bed radiation therapy. Two hundred eighty-six patients underwent a median 66 Gy to the prostate bed between 1998 and 2010. Daily IGRT using either visible surgical clips or bony KV matching was initiated institution-wide in 2007. Cases were reviewed for clinical and dosimetric detail, including manual measurement of CTV to PTV expansion. Biochemical recurrence was defined as a rising PSA >0.2 ng/mL. Toxicity was graded according to CTCAE v 4.0 and RTOG/LENT late morbidity scales. Statistical analysis was performed using STAT v. 11 (two-sided α = 0.05). Actuarial late morbidity and biochemical progression-free survival (BPFS) are reported at 2 years. One hundred seventy-three men were treated without daily imaging. Of the 113 with IGRT, 64 were matched to radio-opaque surgical clips, with the remainder matching to bone. CTV-PTV expansion varied between IGRT and non-IGRT (7 v 15 mm, p < 0.001), but not by type of IGRT matching (p = 0.85). Statistical differences in median follow-up (21 vs 49 months), rates of SVI (30% vs 17%), and Gleason score (IGRT higher) existed (all p < 0.02), with no difference in other pre-operative, post-operative, or radiation factors. Acute toxicity was significantly reduced by the use of IGRT (Table). However, there was no benefit to IGRT for late GI or GU toxicity; in fact, the IGRT group had slightly higher Grade 2+ morbidity rates (Table). Patients treated with IGRT had worse BPFS at two years (63% vs 86%, p < 0.001). This difference persisted on multivariate analysis which included age, ADT, GS, pT, surgical margin status, adjuvant/salvage, use of 3D/IMRT, and type of IGRT matching, with a hazard ratio of 2.51 (95% CI, 1.57-4.02, p < 0.001). In a single institution retrospective series, use of IGRT was associated with an increased risk of biochemical failure independent of other clinical and treatment parameters. We hypothesize that this is due to the reduction in CTV-PTV expansion with inadequate compensation for CTV localization and deformation by image guidance.Poster Viewing Abstract 2417; TablePORT outcome based on use of IGRTIGRTNon-daily imagingPAcute Gr2+ GI/GU morbidity2%/4%28%/16%<.001/.0022-yr late Gr2+ GI/GU morbidity7%/39%4%/31%.30/.032-yr BPFS (95% CI)63% (51-72%)86% (79-90%)<.001 Open table in a new tab