Explore the Scientific R&D Platform

Try Now
Pricing
MENU
Try Now
Pricing

Sample Size Determination Example

AREA: Simon's Two Stage Phase II Design | Oncology

nQuery In Use Background Web  Opacity 35
Simon's Two Stage Phase II Design Worked Examples

Title:
A phase II trial of Sunitinib malate in recurrent and refractory ovarian, fallopian tube and peritoneal carcinoma

Objective: To assess the efficacy and safety of Sunitinib® in recurrent ovarian, fallopian tube and peritoneal carcinoma.

Year: 2013

Source: Gynecologic Oncology, Volume 128, Issue 2

Link: https://www.sciencedirect.com/science/article/pii/S009082581200652X

Protocol: https://clinicaltrials.gov/ct2/show/NCT00768144

Clinical Area: Oncology

Sample Size Section in Paper/Protocol:

“This study used a two-stage design to evaluate efficacy of sunitinib based on overall
response (OR) defined as complete response (CR) or partial response (PR). The null and
alternative OR rates were 5% and 20%.
If one or more patients enrolled in the stage one cohort (n=17 patients) achieved PR or better than accrual would proceed to stage two (n=18 patients). There was 42% probability of stopping the trial at stage one if the true OR rate was 5%. With 35 patients, this design had 85% power to detect the 15% difference in OR rates assuming 2-sided type I error rate of 0.05 .”


Summary of Necessary Parameter Estimates for Sample Size Calculation:

 Parameter Value
Significance Level, ⍺ 0.05
Maximum Ineffective Proportion, 𝜋 0 0.05
Minimum Efficacy Proportion, 𝜋 1 0.2
Power 85%
Design Criterion Minimax (Note: This was not
explicitly stated in the protocol)

 

Email me this example in PDF format

Step 1:
Select the POT11 Two Stage Phase II Design (Simon’s Design) table from the Select Test window.

This can be done using the radio buttons or alternatively, you can use the search bar at the end of the Select Test window.

Two Stage Phase II (Simons Design) -  Sample Size Example 2_Page_3_Image_0001

Step 2:
Enter the parameter values for sample size calculation taken from the study design. The required sample size will automatically calculate once values for ⍺, 𝜋 0 , 𝜋 1 and Power are entered. (Note that the “Optimum Search Value” and “Lower N Limit” parameters are only required under the “Optimum” design criterion.)

Two Stage Phase II (Simons Design) -  Sample Size Example 2_Page_4_Image_0001

The analysis results in a total sample size of 35, with 17 patients in stage one, and a further 18
added for stage two. The actual significance level is 0.028 and the power of the test is 85.01%. At the final analysis, if four or less responses are observed in total, then the drug is rejected. The overall probability of early termination is 41.8%.

Step 3:
Once the calculation is completed, nQuery provides an output statement summarizing the results.

It states:
“Under the minimax design criterion, a sample size of 35 is required to test a null hypothesis of H₀: π ≤ 0.05 versus an alternative hypothesis of H₁: π ≥ 0.2 with a one-sided significance level of 0.028 and 85.01% power, where π is the true proportion of successes. This design results in an expected sample size of 27.474 and a probability of early termination of 0.418.

If the number of responses is less than or equal to 0 out of 17 subjects in the first stage then the trial will be stopped. If the trial proceeds to the second stage, 35 subjects in total will be studied. If 4 or less responses are observed, then the drug or treatment is rejected.”

Two Stage Phase II (Simons Design) -  Sample Size Example 2_Page_4_Image_0002

Step 4:
We can compare the Minimax design with the Optimum and Single Stage designs by completing two additional columns in nQuery with the same ⍺, 𝜋 0 , 𝜋 1 and power parameters and selecting the appropriate Design Criterion.

Two Stage Phase II (Simons Design) -  Sample Size Example 2_Page_5_Image_0001

In this case, the optimum design results in a total sample size of 41, while the single-stage and
minimax designs result in a total sample size of 35.

Step 5:
nQuery advanced also provides plotting options. To access the plotting tools, highlight the
completed columns that you wish to work with, go to the menu and select:
Plot > User-Selected Rows.

Two Stage Phase II (Simons Design) -  Sample Size Example 2_Page_5_Image_0002

In this case, we will display how the total sample size for each design criterion is affected when
the maximum ineffective proportion varies between 0.005 and 0.075

Two Stage Phase II (Simons Design) -  Sample Size Example 2_Page_6_Image_0001

The Edit button at the top of the output allows users to customize the appearance of the plot.

Two Stage Phase II (Simons Design) -  Sample Size Example 2_Page_6_Image_0002

The null hypothesis (response = 5%) reflected a response rate of no clinical benefit while the
alternative hypothesis (response = 20%) reflected a desired response. The type 1 error (falsely
accepting a non-promising therapy) rate was set at 0.05 and the type II error (falsely rejecting a promising therapy) was set at 0.15. A two-stage design using a two-side type 1 error rate of
0.05 was conducted enrolling 17 evaluable participants in stage one and 18 participants in
stage two if one or more participants achieved a CR or PR

Email me this example in PDF format

Free 14 Day Trial

Demo Icon white
No download required