Phase II Clinical Trials

Question 1

What is the main reason why phase II trials are performed?

Answer 1

As a proof of concept that there is some preliminary efficacy from the therapy in question.

Question 2

What is the outcome measure often used in solid tumours?

Answer 2

RECIST

Question 3

What kind of outcome measures are used in phase II trials?

Answer 3

Short term, measurable outcomes such as tumour response.

Question 4

How does a phase II trial help in the eventual design of a phase III trial?

Answer 4

Gives preliminary information on how the treatment works. Also gives information on the logistical aspects like patient recruitment in a specific disease population. This information will aid in forming a well designed phase III trial subsequently.

Question 5

How is random error reduced?

Answer 5

Using a larger sample size.

Question 6

What is the cause of random error?

Answer 6

Due to natural variation in the way patients respond. No two populations of patients will be the same.

Question 7

In what scenarios are phase II trials most effective?

Answer 7

When the cancer is rare and it takes a long time to recruit participants.
When there are few treatments providing benefit for patients

Question 8

When was a phase II trial sufficient for the approval of an oncology drug?

Answer 8

Gefitinib for EGFR positive NSCLC

Question 9

What role does toxicity have in phase II trials?

Answer 9

It’s a key outcome measure in deciding whether or not to proceed to phase III. Many trials have a pre-defined stopping rule e.g a number of severe treatment related adverse events that would cause the trial to stop early.

Question 10

What is the Fleming’s single stage design?

Answer 10

A single group of patients is studied and all receive the experimental treatment. The results observed are compared with the results seen in historical control data. A decision is made to continue once a target clinical effect is achieved, otherwise the trial is stopped.

Question 11

What is the decision rule in Flemings trial design?

Answer 11

number of responders out of all the patients enrolled needed to proceed into a phase III trial.

Question 12

What is Simon’s two stage design?

Answer 12

A small number of patients are recruited for the first stage. If by the end of the first stage the response looks promising, continue onto the second stage, recruiting a larger number of patients. At the end of stage 2, look at all the data from both stages and determine whether there is sufficient evidence to support a phase III trial.

Question 13

What are the strngths of Simon’s two stage design

Answer 13

Limits exposure of an unsuccessful treatment. It also saves time and money if it doesn’t get beyond the first stage.

Question 14

What are the limitations of Simon’s two stage design?

Answer 14

You have two chances of making the wrong decision. It is non-randomized, computer software is needed to calculate sample sizes.

Question 15

What is the advantage of PFS compared to OS?

Answer 15

PFS events are quicker (progression occurs before death). Treatment effect in PFS is often larger than OS as there is often a delay in progression due to the treatment, but not necessarily in the time to death/OS.

Question 16

What is a complete response?

Answer 16

Disappearance of all target and non-target lesions and no new lesions for at least 4 weeks

Question 17

What is a partial response?

Answer 17

Larger than 30% reduction in the sum of longest diameters of target lesions (compared to baseline) for at least 4 weeks.
No progression of non-targeted lesions
No new lesions

Question 18

What is progressive disease?

Answer 18

At least 20% increase in sum of longest diameters of target lesions, relative to the smallest sum since the start of treatment.
Progression of non-targeted lesions
New lesions

Question 19

What is stable disease?

Answer 19

Neither sufficient tumour shrinkage to qualify for partial response nor sufficient increase to qualify for progressive disease

Question 20

What are the target lesions that are measured in order to observe tumour response?

Answer 20

Up to five lesions chosen
Max 2 per organ
Solid tumour lesions are measured along the longest diameter and must be at least 10mm long
Pathological lymph nodes are measured perpendicular to the longest diameter and must be at least 15mm wide.
The size of all lesions are summed so tumour response can be calculated

Question 21

If a patient is both a partial responder, then experiences a complete response; which response is measured in the trial endpoint?

Answer 21

The best response the patient experiences - in this case, a complete response.

Question 22

What are the problems associated with response as an endpoint>

Answer 22

It may not correlate with survival in phase III where OS and PFS are used as primary endpoints.
It assumes the tumour is measurable, might not be the case
Because disease progression could be rapid -so would require a number of expensive scans to be done in order to measure the response which isn’t achievable.

Question 23

What is the problem with time to event as an endpoint?

Answer 23

The real time to event is not accurately measured because it is only seen at a scheduled scan.

Question 24

What is the purpose of a phase II study?

Answer 24

To obtain preliminary evidence on whether a new treatment might be effective (whether it can influence a clinically relevant outcome measure such as mortality or disease progression). Safety is still monitored closely. The results of a phase II often help in the design of a phase III trial.

Question 25

Why are phase II studies sometimes referred to as pilot studies?

Answer 25

Because they are used to assess if a phase III trial is likely to be successful. The study is designed and conducted in a similar manner to a phase III trial, but the protocol specifies that an early assessment is made after a proportion of subjects have been recruited (e.g 25%). Or the trial has run for a fixed length of time.

Question 26

What issues are often raised by pilot studies?

Answer 26

Examining the proportion of eligible subjects approached who agree to participate (acceptance or uptake rate). If accrual is low, what might be the reason for this?

Question 27

What are the advantages of a single arm study?

Answer 27

All resources are concentrated on one group. A predetermined response rate might be used in order to justify further investigation

Question 28

How could a predetermined response rate in a single arm study be used to justify further investigation?

Answer 28

Say a current treatment has a response rate of 20%, and the expected response rate of a new drug is 35%, the sample size would be 56 patients. Of which, 17 or more would be required to respond to indicate that the true response rate is greater than 20%. This defines a cut-off point in which further investigation is justified or not.

Question 29

In a single arm trial of 56 patients, why are 17 or more patients required to indicate that the true response rate is greater than 20% (given that 17/56 is 30%.

Answer 29

Because the 95% confidence interval (one-sided) has a lower limit of 20.4%. The lower limit excludes the possibility of a true underlying rate of 20% with sufficient certainty. i.e the new treatment response rate is likely to be greater than 20%

Question 30

What is the stopping rule?

Answer 30

In a single arm, two stage trial design; the intervention is trialed on a small number of patients, and they are assessed at the end of this stage. If a certain number respond, the trial continues and a second group of subjects are recruited. Otherwise the trial stops.

Question 31

When is the single arm two stage trial design used?

Answer 31

When the outcome is based on counting people. They are useful when the new intervention might have serious side effects or is expensive, because only a few subjects are given the intervention, which may have no true benefit.

Question 32

What are limitations of the single arm two stage design?

Answer 32

After the first stage is reached, centres probably need to stop recruiting until the initial assessment is made and then need to start re-recruiting if enough patients respond.
Also the decision to go to the second stage relies on the response of only one or two subjects. This could lead to an effective drug not being investigated further, or the opposite, an ineffective drug is investigated because a sufficient number of patients happened to show a response.

Question 33

What are the benefits of a randomised phase II study in terms of eventual phase III study design?

Answer 33

The results found in each arm are used to design the corresponding arms in phase III, in particular determining sample size. It also provides information on recruitment rates, subjects willingness to participate in a randomised study and possible logistical problems, all of which help in future studies/

Question 34

What does the sample size method assume to be known with certainty when producing a sample size?

Answer 34

The effect expected from current treatment/ the standard of care being compared against.

Question 35

What significance level is commonly used in phase II trials?

Answer 35

5% - the probability that the trial finds a predefined difference between two treatment arms by chance.

Question 36

What information is needed to calculate sample size in phase II trials?

Answer 36

The expected effect in the new intervention group
The effect in patients given standard treatments (assumed to be known with certainty) 
Significance level (usually 5% at the one sided level)
Power (usually 80 - 90%)

Question 37

When is the survival rate at a critical time point more appropriate to use instead of median survival?

Answer 37

If the median survival has not been reached or it is too dependent on one or two events

Question 38

Why might a phase II trial produce an over-estimation on a treatments effect?

Answer 38

Because they are usually conducted in a few specialist centres and by experienced healthcare professionals. An observed treatment benefit, especially if the trial is not blind may not be found in routine practice.