Performance status restriction in phase III cancer clinical trials.

2059 Background: Patients with good performance status (PS) tend to be favored in randomized clinical trials (RCTs), possibly limiting the generalizability of trial findings. We sought to characterize trial-related factors associated with the use of eligibility criteria that restrict patients by PS, and analyze patient accrual breakdown by PS. Methods: We searched ClinicalTrials.gov for phase III RCTs between 2003-2018. Randomized multi-arm trials assessing a therapeutic intervention in cancer patients were included. PS data were extracted from corresponding manuscripts. Trials with PS restriction Eastern Cooperative Oncology Group (ECOG) ≤1 were identified. Factors associated with PS restriction were determined, and trial patient accrual was analyzed. Results: Six-hundred trials were included with PS data for 238,213 patients. In total, 527 studies (87.8%) specified an upper PS restriction cutoff as part of their exclusion criteria, and 237 studies (39.5%) had a strict inclusion criterion of patients with ECOG PS ≤1. Enrollment criteria restrictions based on PS (ECOG PS ≤1) were more common among industry-supported trials (P< 0.001) and lung cancer trials (P < 0.001). Nearly half of trials that led to subsequent FDA approval included strict PS restrictions. Binary logistic regression revealed stable use of restrictive PS eligibility criteria between 2007-2018 (P= 0.789). The vast majority of patients enrolled across all trials had an ECOG PS of 0 to 1 (96.3%). Even among trials that allowed patients with ECOG PS ≥2, only 8.1% of enrolled patients had a poor PS (ECOG 2 or higher).Trials of hematologic cancers had the largest proportion of patients with ECOG PS ≥2 (8.7%), while lung, breast, gastrointestinal and genitourinary trials all included less than 5% of patients with poor PS (P< 0.001). Only 4.8% of patients enrolled in trials that led to subsequent FDA approval had a poor PS. Conclusions: The use of PS restrictions in oncologic RCTs is pervasive, and exceedingly few patients with poor PS are enrolled. The selective accrual of healthier patients has the potential to severely limit and bias trial results. Future trials should consider a wider cancer population with close toxicity monitoring, to ensure generalizability of results, while maintaining patient safety.

Modelling the patient accrual with truncated Gaussian mixture distribution for the accurate estimation of sample size in survival trials

In survival trials with fixed trial length, the patient accrual rate has a significant impact on the sample size estimation or equivalently, on the power of trials. A larger sample size is required for the staggered patient entry. During enrollment, the patient accrual rate changes with the recruitment publicity effect, disease incidence and many other factors and fluctuations of the accrual rate occur frequently. However, the existing accrual models are either over-simplified for the constant rate assumption or complicated in calculation for the subdivision of the accrual period. A more flexible accrual model is required to represent the fluctuant patient accrual rate for accurate sample size estimation. In this paper, inspired by the flexibility of the Gaussian mixture distribution in approximating continuous densities, we propose the truncated Gaussian mixture distribution accrual model to represent different variations of accrual rate by different parameter configurations. The sample size calculation formula and the parameter setting of the proposed accrual model are discussed further.

Time-to-event model-assisted designs for dose-finding trials with delayed toxicity

Summary Two useful strategies to speed up drug development are to increase the patient accrual rate and use novel adaptive designs. Unfortunately, these two strategies often conflict when the evaluation of the outcome cannot keep pace with the patient accrual rate and thus the interim data cannot be observed in time to make adaptive decisions. A similar logistic difficulty arises when the outcome is late-onset. Based on a novel formulation and approximation of the likelihood of the observed data, we propose a general methodology for model-assisted designs to handle toxicity data that are pending due to fast accrual or late-onset toxicity and facilitate seamless decision making in phase I dose-finding trials. The proposed time-to-event model-assisted designs consider each dose separately and the dose-escalation/de-escalation rules can be tabulated before the trial begins, which greatly simplifies trial conduct in practice compared to that under existing methods. We show that the proposed designs have desirable finite and large-sample properties and yield performance that is comparable to that of more complicated model-based designs. We provide user-friendly software for implementing the designs.

Conducting clinical trials—costs, impacts, and the value of clinical trials networks: A scoping review

Background A significant barrier to conducting clinical trials is their high cost, which is driven primarily by the time and resources required to activate trials and reach accrual targets. The high cost of running trials has a substantial impact on their long-term feasibility and the type of clinical research undertaken. Methods A scoping review of the empirical literature on the costs associated with conducting clinical trials was undertaken for the years 2001–2015. Five reference databases were consulted to elicit how trials costs are presented in the literature. A review instrument was developed to extract the content of in-scope papers. Findings were characterized by date and place of publication, clinical disease area, and network/cooperative group designation, when specified. Costs were captured and grouped by patient accrual and management, infrastructure, and the opportunity costs associated with industry funding for trials research. Cost impacts on translational research and health systems were also captured, as were recommendations to reduce trial expenditures. Since articles often cited multiple costs, multiple cost coding was used during data extraction to capture the range and frequency of costs. Results A total of 288 empirical articles were included. The distribution of reported costs was: patient management and accrual costs (132 articles), infrastructure costs (118 articles) and the opportunity costs of industry sponsorship (72 articles). 221 articles reported on the impact of undertaking costly trials on translational research and health systems; of these, the most frequently reported consequences were to research integrity (52% of articles), research capacity (36% of articles) and running low-value trials (34% of articles). 254 articles provided recommendations to reduce trial costs; of these, the most frequently reported recommendations related to improvements in: operational efficiencies (33% of articles); patient accrual (24% of articles); funding for trials and transparency in trials reporting (18% of articles, each). Conclusion Key findings from the review are: 1) delayed trial activation has costs to budgets and research; 2) poor accrual leads to low-value trials and wasted resources; 3) the pharmaceutical industry can be a pragmatic, if problematic, partner in clinical research; 4) organizational know-how and successful research collaboration are benefits of network/cooperative groups; and 5) there are spillover benefits of clinical trials to healthcare systems, including better health outcomes, enhanced research capacity, and drug cost avoidance. There is a need for more economic evaluations of the benefits of clinical research, such as health system use (or avoidance) and health outcomes in cities and health authorities with institutions that conduct clinical research, to demonstrate the affordability of clinical trials, despite their high cost.

Impact of clinic characteristics on the adoption of a guideline-based standing order algorithm and patient accrual in the pragmatic cluster-randomized trial SWOG S1415CD (NCT02728596).

60 Background: The S1415CD intervention requires the integration of guideline-based prescribing recommendations and standing orders for primary prophylactic colony stimulating factors into existing chemotherapy order systems at community practices within the National Cancer Institute’s Community Oncology Research Program. We looked at the impact of clinic level characteristics on the length of time needed to successfully adopt the intervention and subsequent patient accrual. Methods: We calculated the length of time between randomization and intervention completion for each intervention arm clinic and classified them as short onset (2-5 months, N = 5), medium onset (6-8 months, N = 12) or long onset (10-12 months, N = 7). We compared baseline survey responses about clinic characteristics to onset times. Results: Type of EMR software and the number of chemotherapy regimens reconfigured for the trial had no effect on onset time. All short and medium onset clinics placed orders through an EMR, while 5 of 7 long onset clinics used paper orders. Long onset clinics had less reported nurse involvement in the reconfiguration workflow (change initiation, approval, fulfillment and dissemination) at 14% of clinics vs. 25% of medium onset and 75% of short onset clinics. The average weekly patient accrual rates observed after intervention completion were 1.0 in the short onset (range 0.6-1.5), 0.8 in the medium onset (0.2-1.3) and 0.6 in the long onset (0.1-2.0). Conclusions: When recruiting clinics for trials that require health record system changes, it may be helpful to consider aspects of the system modification workflow such as type of hospital departments involved, as clinics with less nursing involvement may take longer to complete the changes. The inclusion of clinics using different EMR software did not impede onset, but clinics using paper may require more time. Length of onset had no meaningful impact on weekly accrual rates; however, it did determine when clinics could start recruitment, affecting the total number of months clinics could recruit during the study accrual period. Clinical trial information: NCT02728596.