Single-arm Study(Trial)

Randomized and Single-Arm Trials

 

An arm of a clinical trial is a group of patients receiving a specific treatment (or no treatment). Trials involving several arms, or randomized trials, treat randomly-selected groups of patients with different therapies in order to compare their medical outcomes. Experimental arms, which receive an experimental drug, are compared with control arms, which can receive an active comparator (another therapy used to treat the same condition as the experimental therapy), a placebo comparator (an inactive therapy), a sham comparator (an inactive therapy made to look identical to the active therapy), or no intervention. Some clinical trial designs, such as the lung cancer master protocol design, allow for several experimental drugs to be tested simultaneously.

Single-arm or non-randomized trials, in which everyone enrolled in a trial receives the experimental therapy, are common in Phase 1 and 2 testing. While Phase 3 trials are frequently randomized in order to provide more precise data on safety and effectiveness outcomes, a Phase 3 trial might use a single arm if a small target population makes conducting a randomized trial impractical. In these cases, researchers must use endpoints like response rate that can demonstrate clinical impact in the absence of direct comparison data. 

 

ref: https://www.focr.org/randomized-and-single-arm-trials

 

1. Introduction

Many structural designs can be considered when planning a clinical trial. Common clinical trial designs include single-arm trials, placebo-controlled trials, crossover trials, factorial trials, noninferiority trials, and designs for validating a diagnostic device. The choice of the structural design depends on the specific research questions of interest, characteristics of the disease and therapy, the endpoints, the availability of a control group, and on the availability of funding. I discuss common clinical design structures, highlight their strengths, limitations, and assumptions, and provide guidance regarding when these designs may be considered in practice.

2.1 Single-arm trials

The simplest trial design is a single-arm trial. In this design, a sample of individuals with the targeted medical condition is given the experimental therapy and then followed over time to observe their response. This design is employed when the objective of the trial is to obtain preliminary evidence of the efficacy of the treatment and to collect additional safety data, but is not generally used as confirmation of efficacy. The design may be desirable when the available patient pool is limited and thus it is not optimal to randomize many participants to a control arm.

When designing single-arm trials, it is important to clearly define the goal or hypothesis of interest. For example, in a trial with a binary outcome (e.g., response vs. no response) the goal may be to show “any effect” (i.e., the null hypothesis is “zero response” or equivalently that the lower bound for the confidence interval for the response rate is greater than zero). A “minimum clinically relevant response rate” (r_min) would be identified to size the trial. The trial would be sized such that if the true response was r_min, then the probability of the lower bound of the confidence interval for the response rate being above zero (i.e., rejecting the null hypothesis of “zero response”) was equal to the desired power.

This trial design has several limitations and despite the design simplicity, the interpretation of the trial results can be complicated. First, there is an inability to distinguish between the effect of the treatment, a placebo effect, and the effect of natural history. Responses could theoretically be due to the efficacy of the treatment, a placebo effect of an inefficacious therapy, or to a spontaneous or natural history improvement. For a subject that has responded, it could be argued that the subject would have responded even without treatment or that the subject responded because they thought that they were receiving efficacious therapy. Furthermore, it is also difficult to interpret the response without a frame of reference for comparison. For example, if a trial is conducted and no change in the subject status is observed, then does this imply that the therapy is not helpful? It may be the case that if the subjects were left untreated then their condition would have worsened. In this case the therapy is having a positive effect, but this effect is not observable in a single-arm design.

Due to these limitations, single arm trials are best utilized when the natural history of the disease is well understood when placebo effects are minimal or nonexistent, and when a placebo control is not ethically desirable. Such designs may be considered when spontaneous improvement in participants is not expected, placebo effects are not large, and randomization to a placebo may not be ethical. On the other hand, such designs would not be good choices for trials investigating treatments for chronic pain because of the large placebo effect in these trials.

Despite the limitations, single-arm trials may be the only (or one of few) options for trials evaluating therapies for which placebos are not ethical and options for controlled trials are limited. Single-arm trials have been commonly implemented in oncology. Oncology trials often employ a dose at or near the maximum tolerated dose (MTD, known from Phase I trials) to deliver the maximum effect and thus frequently employ single dose trials. The primary endpoint is often tumor response, frequently defined as a percentage decrease in tumor size. Evans et.al. (Evans et al 2002) describes a Phase II trial evaluating low-dose oral etoposide for the treatment of relapsed or progressed AIDS-related Kaposi’s sarcoma after systemic chemotherapy. The primary objective of the trial was to estimate the objective tumor response rate. A response was defined as at least a 50% decrease in the number or size of existing lesions without the development of new lesions. A two-stage design was employed with the plan for enrolling 41 total subjects. However if there were no objective responses after the first 14 subjects have been evaluated, then the trial would be discontinued for futility, noting that if the true response rate was at least 20%, then the probability of observing zero response in the first 14 subjects is less than 0.05. Notably, responses were observed in the first 14 subjects, the trial continued, and etoposide was shown to be effective. Recently the FDA also granted accelerated approval of ofatumumab for the treatment of chronic lymphocytic leukemia refractory to fludarabine and alemtuzumab based on the results of a single-arm trial.

 

ref: Scott 2010 J Exp Stroke Transl Med. 2010 Feb 9; 3(1): 8–18.