Seven reproductive toxicity criteria will be considered with equal weight in order to promote consistency and avoid investigator biases in assessing preclinical reproductive toxicity data, members of FDA's reproductive toxicity working group decided.

Given that investigators are coming from a variety of backgrounds, "there is concern that an individual...who is an expert in pharmacokinetics will put all the weight on the exposure difference, while someone who is a traditional pharmacologist might weight entirely with the pharmacodynamic relationship," FDA's Center for Drug Evaluation & Research Associate Director-Pharmacology and Toxicology Joseph DeGeorge, PhD, commented at a June 24 FDA/Industry workshop on preclinical assessment of reproductive toxicity data.

Pharmacology and Toxicology senior scientist David Morse, PhD, added that after much discussion, the group found that "if we weighted them differently, when we went back and ran another product through the process we came out with a spurious result."

The working group, chaired by Morse, held the public meeting to solicit comments on its integrated risk assessment tool being developed to evaluate "adverse reproductive outcome in humans based on data drawn from multiple species." The working group functions under the reproductive toxicity committee of FDA's Pregnancy Labeling Task Force, which is revising pregnancy labeling categories.

After running a number of compounds through the risk assessment process, SmithKline Beecham Reproductive Toxicology Group Director Patrick Weir said he found the integrated risk assessment tool to be "an extremely well organized weight of evidence method."

The approach evaluates three factors of reproductive toxicity (fertility & fecundity, parturition, lactation) and four developmental toxicities (developmental mortality, dysmorphogenesis, alterations to growth, functional toxicity). If any of the seven endpoints is determined to produce a positive signal, it is then independently evaluated according to six criteria, which the tool breaks down into further elements.

"Assessment within any one of the individual factors should not be an arithmetic summation of the contributory elements," the procedure states, "but an integration made with regard for the quality and nature of the data under consideration."

Based on this method, each signal is assigned a value reflecting the assessed concern: increased (+1), no change (0), or decreased (-1) risk; and those values are tallied to determine overall risk of the endpoint. A net value of over +3 would suggest a significant degree of concern for human reproductive risk and under -3 would suggest no known concern.

"The assignment of concern for each of the six factors of the integration tool reflects a weight-of-evidence assessment taking into account the quality and nature of the data under consideration for each of the contributory elements within the factor," the proposal continues.

The working group determined that assessments should be made according to the investigator's "scientific judgment." Concerned about variation and subjectivity, however, one industry representative recommended that the investigator's assessment should be augmented by the judgment of an "expert."

FDA observed that the algorithm has some objectivity safeguards built into it. "We went through some exercises of looking at the drugs and saying that...' this should be of lower concern or this should be of higher concern'...and actually found it pretty hard to fudge it," FDA's Mark Vogel remarked. "One of the advantages of the system is that you can't take an overall impression and make it work out the way you want it to work out."

FDA Office of Drug Evaluation IV Acting Director and Pregnancy Labeling Task Force Chair Sandra Kweder, MD, acknowledged, however, that when "we do develop an algorithmic approach like this, as an agency we recognize that we may have to make exceptions."

SB's Weir commented that "one big criticism against the existing labeling is 65% of the compounds are in [pregnancy] category C," denoting an adverse effect on animal reproduction, "and what people may be expressing concern for, is now we end up with 85% of the compounds between +2 and -2."

DeGeorge emphasized that the integrated risk assessment method is not intended to translate directly into labeling. "This tried to put a finer line on those compounds which were a 'C'," he said. Vogel added that FDA will be "looking to the +1's and -1's" and asking for narrative descriptions.

Reproductive Toxicity Factors Weighted Equally To Guard Against Bias - FDA

More from Archive

More from Pink Sheet