Eleven patients with histologically confirmed, organ confined (T1-2 N0 M0) adenocarcinoma of the prostate (Gleason score 6–7, PSA levels below 20 ng/ml) were enrolled in a prospective study for evaluation of acute and chronic toxicity of IMRT to 78 Gy to the target volume by using the hydrogel spacer SpaceOAR™ (SpaceOAR™ System, Augmenix Inc., Waltham, MA) for rectal separation. The choice for this PEG-based hydrogel compound was derived from the evaluation of biocompatibility, residence time and costs as discussed by Susil et al. . The prospective study was approved by our institution’s ethics committee (Ethik-Kommission an der Medizinischen Fakultät der Eberhard-Karls-Universität, reference number 079/2011MPG23, study identification number in the German Clinical Trials Register: DRKS00003273). Written informed consent was obtained from all patients. Patients with a high risk of adhesions in the perirectal space, e.g. suffering from inflammatory bowel disease, chronic prostatitis and perianal disease or T3-tumors were not eligible.
All patients underwent prostate MRI (magnetic resonance imaging) to exclude extraprostatic spread. The injection of the hydrogel was performed in an outpatient setting using local anaesthesia and oral antibiotic prophylaxis. After transperineal needle insertion between the rectum and the Denonvillier fascia and hydrodissection with saline under ultrasound control, the hydrogel was injected. A subsequent MRI scan was performed to facilitate the radiation planning process by easy visualization of the hydrogel spacer. The distance created between prostate and rectum achieved by the spacer was measured at prostate apex, center and base. To avoid artifacts caused by different filling of seminal vesicles, the prostatic base was defined as prostate 3 mm below the origin of the seminal vesicles.
Radiotherapy was planned on the basis of three subsequent CTs (computed tomography) in the supine position with a slice thickness of 3 mm. The 3 CT datasets were registered with respect to the bony structures using the Treatment Planning Software (TPS) Oncentra Masterplan® (Theranostic GmbH, Solingen, Germany). Image fusion of the post-injection MRI and CT data sets for visualization of the spacer was performed using a mutual information algorithm. Clinical target volumes (CTV) and organs at risk (OAR) were contoured in each of the three CT data sets by two radiation oncologists (ACM, FP) with assistance of a specialized radiologist for prostate cancer (UK). The CTV included prostate only for low risk patients and an additional proximal 1-2 cm of seminal vesicles for intermediate risk patients. OARs comprised rectum extending from the anal verge to the rectosigmoid flexure, entire bladder, large and small bowel if present, bilateral femoral heads, penile bulb and skin.
From the 3 delineated contours for CTV, a single enclosing union was derived to account for interfraction organ motion and volume changes. Expansion of this union by 7 mm isotropically led to the coverage probability planning target volume (PTVCP). Similarly, OAR unions were created from 3 separately delineated contours. The prescribed dose for the PTVCP was 5x2 Gy/week to a total dose of 78 Gy using a coverage probability approach based on an equivalent uniform dose (EUD) concept. The coverage probability approach consists of assigning individual coverage probabilities of the PTVCP and the OARs to each voxel. The cumulative probabilities are then used as local weights in the cost function during IMRT optimization. As described previously, this treatment planning strategy provides robust IMRT plans and optimal rectal sparing in dose-escalated prostate IMRT . Radiation doses to OARs were additionally evaluated by dose-volume-histogram (DVH) parameters.
IMRT treatment plans were generated with the software package Hyperion® (University of Tübingen, Tübingen, Germany) which uses a Monte Carlo dose engine. Serial constraints were implemented for bladder (k=8) and rectum (k=12) to reach a final maximum EUD of 60 Gy and 65 Gy, respectively . Additional dose constraints for rectum were a V70 of 20% and a V75 of 15%, i.e. a percentaged rectal volume (V) receiving the dose of at least 70 or 75 Gy. IMRT treatment was delivered with a 15 MV linear accelerator (Elekta Synergy S, Elekta Oncology Systems®, Crawley, UK) equipped with a 4 mm multileaf collimator in a sliding window technique. The position of the prostate was regularly verified by conebeam CT according to an image-guidance protocol with an online intervention threshold of 3 mm to account for interfractional prostate motion and to monitor filling of rectum and bladder.
Planning CTs and radiotherapy were performed with a bladder-filling protocol and the use of laxatives. Patients with intermediate risk constellation were offered additional antihormonal therapy for 4–6 months. Acute toxicity was documented weekly during radiotherapy and three months thereafter according to RTOG (Radiation Therapy Oncology Group) classification . The statistical analysis was performed with the software package SPSS 19 (SPSS Inc., Chicago, Illinois, USA). Distance between prostate and rectum was compared by the one-sided t-test for dependant variables.