2005D-0340 Guidance for Industry on Acne Vulgaris: Developing Drugs for Treatment
FDA Comment Number : EC3
Submitter : Dr. Todd Plott Date & Time: 12/21/2005 11:12:09
Organization : Medicis Pharmaceutical
Category : Private Industry
Issue Areas/Comments
GENERAL
GENERAL
The following comments are being submitted electronically in 4 parts. The section below is part 2 of 4.

3. The sensitivity and specificity of the IGA against changes in inflammatory lesion counts has not been performed.

At the Advisory Panel Meeting, Dr. Leyden described retrospective validation against inflammatory lesions as inappropriate. He commented that patients with marked clinical improvement in lesion counts may not be `clear or almost clear?, the dichotomized IGA. This would mean that the sensitivity to detect clinical improvement with the dichotomized IGA is not adequate. Dr. Leyden?s proposal was that 2-grade improvement in addition to `clear or almost clear? be used if an IGA is used at all. The Draft Guidance only partially includes this recommendation. If an IGA is used it should be dichotomized based on ?clear, almost clear or 2-grades of improvement. While this accepts the position of Dr. Leyden, the acceptance begs the question of whether any of the proposed IGA success criteria is any better at measuring meaningful clinical improvement.

4. Validation should be prospective.

No prospective validation of the IGA scale against an objective measurement of clinical improvement has been performed.

5. The IGA scale is unnecessary.

The presumption underlying the development of the IGA scale is that there is a problem with a pre-existing measurement or instrument for determining efficacy of a drug product. The pre-existing measurement of efficacy is lesion counting. The objections identified with lesion counting are that it is cumbersome and may not represent clinical improvement. Since it is acne lesions that are the subject of the treatment, direct counts are the most relevant measure of improvement. Thus the objections to use of lesion counts are not an acceptable justification for mandating development of a new untested and untried scale.

6. Well-performing outliers can confound the IGA Scale.

Based on existing data available to the Division, the IGA scale is designed to indicate efficacy based on a small proportion of patients successfully achieving the endpoint of `clear or almost clear?. This may cause the IGA scale to be inaccurate if a small number of patients were to achieve `success? through investigative error or unexpected improvement. If a few patients are incorrectly judged a `success? are in the active arm, a false positive result may occur; if the few successful patients are in the placebo arm, then a `false negative? result may occur. Stated differently, because most patients do not achieve a successful endpoint and therefore do not count towards efficacy, the scale is subject to a high rate of false positive and false negative observations.

7. The IGA scale cannot differentiate products that are effective against one or the other type of lesion.

The point was made at the Advisory Panel that products that are intended for the treatment of only one lesion type, either non-inflammatory lesions or inflammatory lesions, should be approved. The guidance provides for approvals for a single lesion type and requires the use of the IGA Scale even though the scale has been shown to be heavily weighted to improvements in inflammatory lesions.