Single blind randomized Phase III trial to investigate the benefit of a focal lesion ablative microboost in prostate cancer (FLAME-trial): study protocol for a randomized controlled trial

The FLAME-trial is a multicenter randomized controlled trial. Patients are recruited during the intake consultation at the Department of Radiation Oncology in one of the participating centers. After giving informed consent, patients are randomized either to the standard arm or to the experimental arm. Patients in the standard arm receive radiotherapy according to the current standard [4], namely 77 Gy in 35 fractions of 2.2 Gy to the whole prostate. Patients in the experimental arm receive an additional integrated microboost to the macroscopic tumor to a total dose of 95 Gy in 2.7 Gy fractions.

To ensure unbiased assessment of QoL measurements, the patients are blinded to the actual treatment given (receiving an ablative microboost or not). The treating physician needs to be informed about the actual treatment, to be able to judge the treatment plans. This is unlikely to influence the assessment of the objective primary endpoint of the trial.

Men with histological proven intermediate or high risk adenocarcinoma of the prostate, who will receive external beam radiotherapy using optimal position verification with implanted fiducial gold markers, are eligible for the study. Intermediate or high risk is defined according to the currently internationally accepted criteria from Ash et al. [21]. Intermediate risk is defined as patients having one factor of stage T2b-c, or Gleason score = 7, or initial prostate-specific antigen (iPSA) = 10-20 ng/mL. Patients having more than one of these factors or having stage T3, or Gleason score >7, or iPSA > 20 ng/mL are defined as high risk. Patients with low risk tumors are not included in this study, because the treatment outcome for this group is already excellent with a 10-year prostate cancer-specific survival approximating 95% [1].

Exclusion criteria are: previous pelvic irradiation, previous prostatectomy, World Health Organization (WHO) performance score > 2 [22] (= symptomatic, >50% during the day in bed, but not bedbound (score 3) or bedbound (score 4)), International Prostate Symptom Score (IPSS) ≥ 20, transurethral resection of the prostate (TURP) within 3 months from start of the treatment, general contraindications for MRI (i.e. cardiac pacemaker, metal implants or history of severe allergic reaction after administration of contrast agent) or the use of anti-coagulants that cannot be discontinued for the gold markers implantation.

The study protocol is approved by the Medical Ethical Committees of the participating hospitals. Written informed consent will be obtained from all patients.

Randomization is performed by an independent trial center. If a patient meets the inclusion criteria and has provided informed consent, the physician contacts the trial center. To prevent randomly occurring differences in important prognostic factors across the two randomized groups, the randomization is stratified by TURP, hormonal treatment and by centre. A TURP prior to radiotherapy is associated with significantly more late genitourinary toxicity [23, 24]. The likely mechanism of increased late toxicity is related to the relative devascularisation of the urethra after TURP and the decreased capability of the mucosa to repair sublethal damage after radiotherapy [24]. Hormonal treatment is a prognostic unfavorable factor for late genitourinary side effects [23, 25] and erectile impotence [26‐28]. Furthermore, a protective effect for hormonal treatment is reported for acute gastrointestinal side effects [23, 29, 30]. To prevent small numbers of patients in a particular hospital from all receiving, by chance, the same treatment, the hospital is chosen as one of the factors for stratified randomization as well.

Between October 2009 and October 2010, the first 50 patients were included at the Department of Radiation Oncology of the University Medical Center Utrecht. Based on this accrual and the fact that other participating centers recently started to include patients as well, we expect that the accrual will be completed within 5 years from start.

To minimize the positioning errors during treatment all patients are treated with an online position verification protocol using implanted fiducial gold markers [14, 31, 32]. Radiotherapy will be delivered with advanced radiotherapy techniques to be able to create adequate dose distributions.

A mean dose of 77 Gy in 35 fraction of 2.2 Gy is prescribed to the entire prostate gland [33, 34]. The dose in the part of the PTV overlapping the rectum and bladder is limited to keep the risk of severe gastrointestinal and genitourinary toxicity acceptable.

To provide an accurate delineation of the prostate gland with respect to the surrounding tissues [35, 36], the prostate is delineated on a computed tomography (CT) scan combined with a registered magnetic resonance imaging (MRI) scan. The rectum is contoured from the anus or ischial tuberosities to the rectosigmoid flexure or sacroiliac joints. The bladder is completely outlined from the bladder neck to the dome.

The precise execution of the radiotherapy treatment can differ in each participation center, but the treatment will always meet the above mentioned criteria.

Patients randomized to the experimental arm are treated with the current standard of 77 Gy to the whole prostate and in addition receive an integrated microboost to the macroscopic tumor to reach a total dose of 95 Gy in 35 fractions of 2.7 Gy. To delineate the macroscopic tumor within the prostate, different MR imaging techniques are used. In addition to an anatomic T2 weighted sequence, a combination of the following functional imaging modalities can be used [37, 38]. Dynamic contrast-enhanced (DCE)-MRI gives a characterization of the tissue vasculature [39]. With this technique it is possible to detect areas with macroscopic tumor, because tumors tend to contain higher density of leaky blood vessels. With diffusion-weighted imaging (DWI)-MRI the mobility of water molecules is measured [40]. Tumor tissue can be identified on DWI-MRI, because in tumor the extracellular volume is reduced, leading to reduced water diffusion in tumor tissue. MR spectroscopic imaging (MRS) provides metabolic information with cancer regions showing higher choline and lower citrate levels [41].

To evaluate whether the addition of an ablative microboost to the macroscopic tumor within the prostate increases the five-year freedom from biochemical failure rate compared to the current standard of care. Biochemical failure is defined according to the Phoenix definition as a PSA rise of 2 ng/mL above the nadir PSA level [42]. The PSA level will be measured every six months until 10 years after treatment.

Secondary endpoints are treatment-related toxicity, QoL and disease-specific survival. Treatment-related toxicity is measured by the Common Toxicity Criteria for adverse events version 3.0 (CTCAE) [43]. The following adverse events are scored: urinary frequency/urgency, urinary retention, bladder spasms, urinary incontinence, genitourinary hemorrhage, dysuria, rectal or perirectal pain, proctitis, diarrhea, flatulence hemorrhoids, anal incontinence, erectile dysfunction. The physician in attendance scores the complaints before treatment, acute toxicity (weekly during treatment and 4 weeks after treatment) and late toxicity (every six months until 10 years after treatment). All symptoms are registered even if they occur only on one single occasion. Grade > 2 is considered severe toxicity.General health-related QoL is measured using the RAND-36 generic health survey [44], cancer-specific QoL using the European Organization for Research and Treatment of Cancer (EORTC) core questionnaire (QLQ-C30) [45], and the prostate tumor-specific QoL using the EORTC prostate cancer module (QLQ-PR25) [46]. The disease- and treatment-related side effects are only a component of health-related quality of life [47]. To address the components of overall well-being, a general instrument is used in addition to the disease-specific questionnaires [48‐50]. The RAND-36 assesses physical and social functioning, physical and emotional role restriction, mental health, vitality, pain, general health and change in health. The EORTC QLQ-C30 contains five functional scales, three symptoms scales, a global QoL scale and six single-items. The EORTC QLQ-PR25 assesses urinary, bowel and sexual symptoms and functioning, and the side effects of hormonal treatment. The first questionnaire is handed over to the patient one week before treatment at the Department of Radiation Oncology and the next questionnaires are sent to the patient every six months until 10 years after the completion of the treatment. It is important to measure QoL every 6 months after treatment, to be able to determine the point in time at which the QoL changes, for example due to side-effects after treatment.

For disease-specific survival, death with metastases is considered a death caused by the disease.

An independent data safety monitoring board (DSMB) will evaluate the toxicity and clinical outcome. The DSMB receives an update of the toxicity in total and per treatment arm every 3 months. Serious toxicity, defined as any acute or late toxicity requiring surgical intervention, any grade 4 toxicity, and any not-transient (duration >6 months) late toxicities grade 3, will immediately be reported to the DSMB. An overview of the percentage of biochemical recurrences per treatment arm together with the evaluated number of patients per arm will be sent to the DSMB yearly as well as a statistical comparison of the incidence of serious and less serious toxicities in the two arms. The DSMB decides on stopping or continuing the trial. Exact rules cannot be specified in advance, but an increase in the incidence of toxicity (grade 3-4) with 5% or a smaller, but statistically significant increase, are among the reasons to consider stopping the trial. The DSMB can decide to prematurely stop the trial in case the improvement in the experimental arm is higher than anticipated in the trial design.

The statistical power of the study was calculated for the primary endpoint (five-year freedom from biochemical failure). Based on data of two randomized clinical trials [2, 51] reporting the treatment results of patients treated with a radiation dose equal to the dose of the standard arm of our trial, we expect that the five-year freedom from biochemical failure of the standard arm will be approximately 64%. We expect that an additional ablative microboost to the macroscopic tumor will increase this number with at least 10%. A one-sided log rank survival power analysis shows that the length of follow-up after accrual of the last patient should be 3.5 years to detect a difference of 10% (64% and 74% free from biochemical failure after 5 years, in the control arm and experimental arm, respectively) with a power of 80% at a one-sided 5% significance level. This is under the assumption that the patients enter the study during an accrual period of 5 years, 50% of the enrollment is complete when 70% of the accrual time has past and that 20% of the patients in both arms are lost to follow-up during the follow-up period of 5 years. The reason to choose a one sided p-value is that, although extremely improbable, an increase in biochemical failure in the experimental arm would lead to the same action as no difference at all between the two treatment arms. This is because the experimental treatment will only be implemented if it is significantly better than the usual treatment, due to the increased toxicity risk in the experimental arm.

All analysis will be performed according to the intention-to-treat principle. Cox proportional hazards regression will be used to analyze differences in biochemical failure and disease-specific survival between the two treatment arms. Survival curves will be estimated by the Kaplan-Meier technique. Differences between both groups in the incidence of acute side-effects will be tested with the Chi square test. The incidence of late toxicity will be analyzed actuarially with the Kaplan-Meier method, the log rank test and Cox regression analysis. To analyze differences in QoL between the two treatment groups over the time points, a general linear model repeated-measures analysis of covariance will be performed [52]. A change of 10% (or in general, 0.5 standard deviation) of the scale width is perceptible to patients as a meaningful change [53]. Because of the multiple comparisons for the QoL items, the p-value is set at a conservative 0.01 for determining statistical significance [54]. For all other analyses, which do not include the QoL measures, a p-value of < 0.05 is considered statistically significant.