Stereotactic Radiotherapy vs Best Supportive Care in Unfit Pancreatic Cancer Patients

This study is a prospective multicenter randomized controlled trial according to the 'cohort multiple randomized controlled trial (cmRCT)'-design. This study will be nested within the existing prospective observational PACAP cohort (the Dutch Pancreatic Cancer Project; PACAP), in which all 48 centers of the Dutch Pancreatic Cancer Group (DPCG) participate. PACAP-participants have provided informed consent for being randomized in cmRCTs. Patients who are randomized to the interventional arm will be approached for informed consent. Patients who are randomized in the control arm will not be approached as they already provided informed consent(12) .

PACAP(13) is a nationwide registration outcome project including the Dutch Pancreatic Cancer Audit, a nationwide Expert Panel and the Dutch Pancreas Biobank (PancreasParel). The PACAP-cohort is designed according to the cohort multiple randomized controlled trial (cmRCT) design, meaning a large observational cohort of patients, which can be used as a multiple trials facility. Participants in the PACAP-cohort have provided general informed consent for collection of specific data on demographics, quality of life and clinical, laboratory and radiological findings during follow-up after randomization.

To investigate whether SABR relieves tumor-related symptoms such as pain, improves the quality of life and prolongs survival compared to best supportive care, randomization is performed within PACAP-participants with primary pathology-proven non-metastasized PC for whom currently no treatment is available other than palliative care, or those who choose to refrain from radical treatment.

Population (base)

In current clinical practice, there remains a substantial group of patients with non- metastasized PDAC who are medically inoperable based on compromised physiological reserve and/or comorbidities, for whom currently no treatment is available other than best supportive care, or who choose to refrain from radical treatment in the form of surgery or chemotherapy. According to NCR data, this group consists of approximately 300 patients per year. These patients may, however be fit enough to undergo a short course of SABR.

After discussion in the MDT with a surgeon, gastroenterologist, medical oncologist, radiologist, pathologist and radiation oncologist, all specialized in the treatment of PDAC, the decision to withhold radical treatment with surgery and/or chemotherapy is made in frail or elderly patients. If these patients are participants in the PACAP-cohort, they will be followed and treated according to the current practice, i.e. best supportive care. Subsequent randomization for SABR is performed within the subgroup of patients eligible for the PANCOSAR study. This subgroup exists of all PACAP-participants with histologically confirmed primary non-metastasized PDAC for whom currently no treatment is available other than best supportive care, or who refuse radical treatment. In case we find that pathology is lacking in more than seven patients among the first ten patients that were considered to be eligible for this study, an amendment for this initial protocol will be considered. Only patients randomized for the intervention-arm will be approached for participation; patients randomized for the control arm will have their follow-up according to the standard of care. After having obtained the extra IC, patients will be formally included. This concept is further explained in section 6.2.

With three large volume centers (Amsterdam UMC, Erasmus MC Cancer Institute, Rotterdam and UMC Utrecht) the planned number of patients can be recruited from the defined source population in the suggested 2.5 years' time period. All other DPCG centers can participate in the project but patients are only treated with SABR at these three centers.

When calculating the sample size the assumption is that at 6 months, 20% of patients managed with best supportive care will be alive versus 50% of patients treated with SABR. Taking into account that an estimated 30% of the people randomized for the intervention will refuse treatment, dilutes the survival in the intervention arm to 41% (0.7x50% +0.3x20%). Since the control arm is not informed about the intervention, we assume 100% compliance in the control arm.

With alpha 15% (14) and power 80% a sample size of 49 patients per treatment arm is required in this Phase 2 randomized trial. If however the cmRCT design results in a dilution over 20% a different approach of inclusion will be considered.

Stratification will be performed for patients with localized PDAC who refuse surgery and chemotherapy because these patients may present with a potentially better prognosis.

The main analysis will be based on all randomized patients who have provided informed consent for participating in this trial. Because of the PACAP/cmRCT which is explained in detail in section 6.2, we will have to randomize additional patients in order to have informed consent of 49 patients for the intervention arm. In order to analyze treatment effect an additional per protocol analysis will be performed. Time to any event (death, development of metastases or local failure although no study specific imaging is planned) will be calculated from randomization to the documentation of the event of interest. The primary endpoint of survival at six months will be compared between both treatment arms using log-rank analysis and Kaplan-Meier curves.

With regard to the first secondary endpoint, time to decreased global QoL is defined as the time since randomization to a decrease of 20 points in the global QoL of patients, as measured by the EORCT C30 questionnaire global health status. At randomization, global health status is estimated to be 70 on average, and we anticipate that 50% of the best-supportive-care group will experience a 20 point decrease in global QoL on average at 2 months after randomization. We aim for a benefit in time to a decrease of global QoL of at least three months. Sample size calculation for the analysis of this secondary endpoint was calculated at 76 patients in total, using an alpha of 0.05 and a power of 80%. This estimated sample size is within the sample size for the primary endpoint (98 patients).

In order to evaluate pain, which is defined as having an NRS ≥4 and/or the usage of morphine-like medication; a decrease is defined as: 1) a decrease of ≥2 points on the NRS scale compared to baseline, or 2) cessation of morphine-like medication.

Patient accrual is expected to start in September 2020. With at least three large-volume centers involved in the study, it is estimated that patient accrual can be completed within 2.5 years

Data Safety Monitoring Board (DSMB) / Safety Committee

An independent data safety and monitoring committee (DSMB) will evaluate the progress of the study and will examine safety parameters. The DSMB will consist of an independent epidemiologist/statistician who is the chairman and an independent radiation oncologist and surgeon. The members of the DSMB will declare possible conflicts of interest. All involved physicians will repetitively be asked to report any potential AE's caused by the study protocol. These AE's will be listed and discussed with the monitoring committee. The monitoring committee can ask for a full report in order to discuss a specific AE. A copy of this report will be sent to the central ethics board and to the involved physicians. All deceased patients will be evaluated by the safety committee for cause of death and possible trial related SAE's. Every death will be reported to the central ethics board and the local ethics board. All DSMB meetings, open and closed, will be documented. The advice(s) of the DSMB will only be sent to the sponsor of the study. Should the sponsor decide not to fully implement the advice of the DSMB, the sponsor will send the advice to the reviewing METC, including a note to substantiate why (part of) the advice of the DSMB will not be followed.

Other reasons for monitoring efficacy might be for futility, checking the assumptions for sample size calculation or whether criteria for early stopping are met.

STATISTICAL ANALYSIS

All clinical-pathological variables mentioned in the inclusion criteria will be described and analyzed according to continuous variables will be summarized with standard statistics including, means, standard deviations, medians and ranges. Categorical variables will be summarized with frequencies. When appropriate, box plots and cross tables will be used for descriptive statistics of continuous and categorical variables, respectively. P-values below 0.05 will be considered significant. All calculations will be generated by statistical package for social sciences software (SPSS). Calculation of the number of patients that will be needed to address our primary endpoint with a power of 80% and a 15% type I error rate is described in section 4.4.

Primary study parameter(s)

The primary outcome measure of the study will be the overall survival at six months in both groups, measured at the time from randomization. Survival endpoints (disease free survival and overall survival) will be analyzed using Kaplan Meier plots and log rank test with additional analyses using Cox proportional hazards modelling in order to adjust for stratification and prognostic variables. All data will be collected in a secured electronic database (Castor). The outcome parameters will be analyzed with appropriate statistical tests using the statistical program SPSS.

Secondary study parameter(s)

Secondary outcome parameters are both early- and subacute toxicity, pain, Ca19.9 and performance score within the first year after treatment. To investigate early and subacute toxicity, signs and symptoms are graded from 1 to 5 according to the CTCAE version 5.0 for each time point following treatment. The pain score will be assessed using the NRS score. A clinically relevant NRS score improvement is defined as decrease of 2 points to baseline. In addition, the lapse of the use of morphine-like regimens is also scored as improvement. Progression of disease (local and distant) is assessed using CT-scans at 3, 6 and 12 months in the intervention arm using the Response Evaluation Criteria in Solid Tumors (RECIST), version 1.1.

After completion of follow-up of all patients, adverse events will be described to provide an overview of the encountered toxicity. Percentages of incidence will be given for each type and grade of complication. The patient-reported outcomes (QoL) will be scored using the EORTC-QoL C30, Pancreatic Cancer module QLQ PAN26 questionnaires. Quality of life analysis of the patients included in the trial will be included as one of the major objectives of this study. Patients who complete the quality-of life questionnaires at baseline and at least once during follow-up can be included in the analysis. The evaluation of time to decreased global QoL is defined as the time since randomization to a decrease of 20 points in the global QoL of patients, as measured by the EORCT C30 questionnaire global health status.

Interim analysis (if applicable)

After inclusion of half of the patients an interim analysis will be performed as described in paragraph 4.4, 7.3 and 7.5.