Alzheimer's disease: Understanding the results of responder analyses

It is important to know whether a change in the score actually has a noticeable effect in patients' everyday lives. Responder analyses can help to answer this question: For each person in the study, the researchers compare the two outcomes of the questionnaires (the scores before and after treatment) and calculate the difference between them. Then they count the number of patients who have changed noticeably – in other words, in ways that clearly affect their everyday life. In the lecanemab study, those people were considered to be "responders."

This question is not always easy to answer. Noticeable changes make an important difference in the patient's life. IQWiG used a threshold derived from the various questionnaires: If a patient's score changed by at least 15%, they were considered to be a responder. This means that they had changed enough for it to be clearly noticeable in everyday life.

In the lecanemab study, people whose symptoms had become noticeably worse were considered to be responders.

Because the CDR-SB score ranges from 0 to 18 points, the threshold here is 2.7 points (15% of 18). So a person is defined as a responder if their score has changed by 3 points or more (only steps of 0.5 points are possible due to the way the questionnaire is evaluated). In the lecanemab study, those people's symptoms had become noticeably worse.

The clinical trial included people with mild cognitive impairment (MCI) as well as people with mild Alzheimer's dementia.

The results of the responder analysis for people with mild cognitive impairment (MCI):

The symptoms became noticeably worse (the CDR-SB score increased by at least 3 points) in

• 22 out of 216 people who had received lecanemab (10.2% of them) and
• 25 out of 217 people who had received a placebo (11.5% of them).

This difference is too small to be able to clearly conclude that it was a result of treatment with lecanemab.

The results for people with mild Alzheimer's dementia:

The symptoms became noticeably worse (the CDR-SB score increased by at least 3 points) in

• 28 out of 103 people who had received lecanemab (27.2% of them) and
• 39 out of 115 people who had received a placebo (33.9% of them).

Here too, the difference is too small to be able to clearly conclude that it was a result of treatment with lecanemab.

The results of the evaluations of other questionnaires (for example, on cognition and quality of life) were similar: The differences between the treatment groups don't allow any clear conclusions to be drawn about the effectiveness of lecanemab.