Prospects for Developmental Immunotoxicity Guidance and an Update on ICH S8

Abstract

Potential adverse effects of drug exposure on the developing immune system had been a relatively neglected area of toxicology until fairly recently. However, with the recent regulatory emphasis on evaluation of drugs for use in pediatric patients, juvenile animal studies have been considered in much more detail than in the past in order to enable clinical trials. Assessment of immunotoxicity potential in these juvenile animal studies has been a natural consequence of drug development for pediatric patients. Unfortunately, the efforts to evaluate developmental immunotoxicity studies have not kept pace with the writing of an immunotoxicology guidance for the International Conference on Harmonisation of Technical Requirements for Registration of Human Pharmaceuticals (ICH). This document has recently reached Step 4 in the approval process: it is thus likely that juvenile animal studies for immunotoxicity would be recommended based on considerations enumerated in the ICH Safety Number 8 (S8) guidance. Sufficient flexibility exists in this document to cover the issue of developmental immunotoxicity. ICH S8 advocates a weight-of-evidence approach, which should be interpreted to indicate that immunotoxicity testing would be conducted based on identified cause(s) for concern rather than as a routine screening method. The presentation reflecting these issues, as presented to attendees of the Pharmaceutical Education Associates workshop on Innovative Methods and Applications for Risk Assessment in Pharmaceutical Development is summarized herein.

Keywords :

• developmental immunotoxicology
• ICH S8

DEVELOPMENTAL IMMUNOTOXICITY

A basic paradigm has arisen within the toxicology community that the developing organism is more vulnerable to xenobiotic insult compared to adults, assumed to be fully developed. This issue has received increased emphasis in recent years, particularly with respect to the developing central nervous system (CNS) and potential exposure to neurotoxicants such as pesticides. However, the issue of developmental neurotoxicity has also become an important consideration in drug development, both because of the increasing use of psychoactive drugs to treat conditions such as hyperactivity/attention deficit disorders and Pediatric Exclusivity, an effort by the U.S. Food and Drug Administration (FDA) to encourage clinical trials in infants, children, and adolescents in order to support labeling claims and instructions for use (The Pediatric Rule, 1998; Rice and Barone, 2000). In particular, the issue of clinical trials has also focused attention on whether drugs developed for use in adults can be safely given to children: thus, what was an effort originally designed to support dosing recommendations on product labels has evolved into a more robust effort to assess safety as well as efficacy in pediatric patients. The integration of developmental neurotoxicity findings into the drug evaluation process is a work in progress, however. For example, methylphenidate has been shown to have effects on the developing nervous and endocrine systems in animals (Greely and Kizer, 1980; Pizzi et al., 1987). The clinical significance of these findings is still debated.

It was a natural progression that toxicologists began to consider the developing immune system as another potential target of adverse drug effects that might not be anticipated based on studies in older animals or adult humans. In fact, there has been a fair amount of research in developmental immunotoxicity, but not nearly to the extent needed to make recommendations concerning either the need for studies or the way in which these studies, when needed, should be conducted (Barnett, 1996; Dietert, 2003). Thus, the two FDA Center for Drug Evaluation and Research (FDA/CDER) guidance documents that address the issue of developmental immunotoxicity do not provide detail on either issue. The U.S. FDA CDER Guidance for Industry on Immunotoxicology Evaluation of Investigational New Drugs (2002) advises sponsors to conduct immunotoxicity assessment as part of standard rodent reproductive/developmental toxicology studies if the drug has been shown to be an immunotoxicant in adults, but limits specific advice to examination of immune system tissues for signs of immunotoxicity. There is no specific recommendation on conduct of immune function assays as part of this evaluation. The draft CDER Guidance for Industry on Nonclinical Safety Evaluation of Pediatric Drug Products likewise advises assessment of immunotoxicity potential of known adult immunotoxicants in an “appropriate” animal model, but does not make specific recommendations.

However, this situation is likely to change significantly in the future. First, publication of the final ICH S8 Guidance on Immunotoxicity Studies for Human Pharmaceuticals lists as one cause for concern that might be taken to indicate the need for specific immunotoxicity testing the specifics of the intended patient population, including age. Thus, it is implied (though not specifically stated) that the developing immune system should be considered as a specific factor in a weight-of-evidence approach to deciding the need for immunotoxicity testing.

Secondly, recent publication of a proposed framework for immunotoxicity testing (Holsapple et al., 2005) presents in detail a practical approach to including immune function testing within traditional Segment III reproductive toxicology studies. The authors advocate an approach designed for hazard identification: thus, exposure to test compound is maximized throughout gestation, lactation, weaning, and early development (in rat pups, the preferred species, this includes direct dosing up to post-natal day 42) followed by gross and histopathology, at a minimum. In addition, a function study such as the T-dependent antibody response assay (TDAR) should be included. Other assays should be considered on a case-by-case basis, and if a signal is observed, additional studies may be conducted to determine if there is a specific “window of vulnerability” for immunotoxic effects.

This is a practical approach that builds on a well-established method in developmental toxicology. The approach differs from traditional rat Segment III studies in that there is direct dosing of pups after lactation, and histopathology has not been part of the standard approach. Neither of these additional steps was considered by the expert panel to create an unreasonable burden. In addition, it can serve as a method that builds on the most robust database in immunotoxicology: that is, the enormous amount of information that has been accumulated using rodents. Although non-rodent models could be considered in certain situations (especially when dealing with protein therapeutics), the default method relies on standard methodology in a comparatively well-understood species with respect to the functional endpoints.

The only truly major issue left unresolved is/are the circumstance(s) under which developmental immunotoxicity studies should be conducted (“triggers”). The primary trigger would be findings of immunotoxicity in standard adult animal toxicology studies. Although other potential triggers are identified, such as structure-activity relationships, intended use (e.g., prevention or perinatal transmission of an infectious disease such as human immunodeficiency virus), or intended neonatal/juvenile exposure, in fact the primary issue is demonstrated immunotoxicity. For example, consider a situation in which a new drug is in a pharmacologic class known to have immunosuppressive activity but does not demonstrate immunotoxic signs in animal studies: should this be considered a “structure activity” issue and immunotoxicity studies conducted anyway? The central conundrum is this: if in fact the developing immune system is more sensitive to insult compared to adults, is it reasonable to use findings in a less-sensitive population to decide the need for immunotoxicity testing? One possible solution to this problem would be to include immune system evaluation as part of Segment III reproductive toxicology studies—this possibility has been discussed (Ladics et al., 2005). Simply put, not enough research/data appears to exist to resolve this issue. However, as discussed next, ICH S8 in fact leaves this question open for regulatory decision making.

ICH S8

On June 14, 2005, a telecom was held in which all parties to ICH agreed to a final text for ICH S8. Although the final text has not been vetted by regulatory counsel at the time of this writing, the Step 4 document that is very likely to be accepted by all ICH parties can be discussed as a complete guidance. The document, when published in final form, will be available at 〈www.fda.gov/cder/guidance〉.

The history of the guidance is long and complicated, so what follows should not be taken as either comprehensive or definitive. Rather, it represents the perspective of the author. Development of guidances on evaluating drugs for immunotoxic potential really began in the early 1990s, and had its origin, at least in FDA/CDER, with the original cohort of drugs intended for treatment of AIDS. It was recognized early in the development of AIDS therapeutics that attention should be given to the possibility that adverse drug effects on immune function would render compounds shown to be effective anti-retrovirals in vitro unsafe for patients. Thus, attempts were made to determine the potential for candidate drugs to produce unintended immunosuppression. There were in fact some early candidate anti-retrovirals that in fact did produce immunotoxic effects (some of which were not appreciated until clinical trials were completed). However, as a general rule, the issue of immunotoxicity did not become as important as anticipated. However, the issue of immunotoxicity and drug development remained important for an unrelated reason: the realization that many adverse drug reactions appeared to be immune-based. That is, the issue of drug allergy became the driving force behind the development of a guidance on immunotoxicity assessment of drugs at FDA/CDER. This is truly paradoxical as will be explained next.

In July 2000, the European Agency for the Evaluation of Medicinal Products (EMEA) Committee for Proprietary Medicinal Products (CPMP) issued a guidance on repeat-dose nonclinical toxicity studies that included an Appendix that has been widely interpreted to mandate testing for immunotoxicity for any new drug for which marketing authority in the European Union (EU) was sought. In October, 2002, FDA/CDER issued its Guidance for Industry on Immunotoxicity Evaluation of Investigational New Drugs that advocated a cause-for-concern approach with respect to determining the need for directed immunotoxicity testing. Thus, two of the three ICH regions had published guidances on drug development that appeared to be in conflict. The Japan Ministry of Health, Labour, and Welfare (MHLW) and Japanese pharmaceutical companies took note of this and proposed that the issue should be resolved within the ICH process. Thus, ICH S8.

It is somewhat ironic that the issue of unintended immunosuppression drove the ICH process on immunotoxicity. In order to facilitate negotiations, a group of Japanese regulators and industry representatives began a survey to determine (1) the extent to which pharmaceutical companies were testing candidate drugs for immunotoxic potential (which, in the following discussion, will be taken as unintended immunosuppression), and (2) whether these studies were revealing adverse immune effects that would not have been anticipated based on routine toxicity studies (RTS). This survey was conducted in two phases, and 39 drugs were eventually found to have had sufficient information provided to make a determination of the contribution of directed immunotoxicity testing to the safety evaluation of drugs in development. The irony is, of course, that the problem that had been identified fairly conclusively, at least within FDA/CDER, was drug allergy, not unintended immunosuppression.

In October 2003, an Expert Working Group (made up of authorities from the six ICH negotiating parties as well as an observer from Canada, and eventually joined by an observer from Switzerland) convened a meeting at the EMEA headquarters in Canary Warf, London, and began examining the results of the Japanese survey. The results of the analysis, which continued in Osaka, Japan in November 2003 and in McLean, Virginia in June 2004, have been submitted and should be consulted when published. However, some simple conclusions were reached based on both this survey as well as other information available to the S8 Expert Working Group (ICH S8 EWG). There were only, at most, four examples where directed immunotoxicity testing could arguably have identified potentially immunotoxic drugs where the effect would have been missed, based on data obtained in RTS. There has been much discussion about whether this is an overestimation of the failure to detect immunotoxicity, but it was the conclusion of the EWG that even if this number was accurate, the results of the survey could not be taken to support directed immunotoxicity testing as a default requirement. Simply put, even in the four examples of “missed immunotoxicity,” the effects observed in the specific immunotoxicity tests were relatively minor and in no case could these findings be shown to have been critical to determining the safety of the drug. Much could be made of the size of the database, the methods chosen for analysis, etc., but the simple fact is the EWG could not find a way to support directed immunotoxicity testing as a default requirement (Weaver et al., 2005).

As has been mentioned, the issue of drug allergy was not addressed. A detailed discussion of this decision will not be attempted here. Simply put: there were no methods that could be recommended (with the single exception of methods to detect ability to produce allergic contact dermatitis—ACD), so there was nothing about which to negotiate. This simple point seems to have been lost on many who have criticized the document. The one issue about which we could have negotiated, basically whether the murine local lymph node assay could be used as a replacement for standard guinea pig assays to detect ACD potential, was determined early in the process to be an issue about which there was no substantial disagreement (all parties accept the LLNA as a stand-alone alternative).

However, it was agreed by the EWG that unintended immunoenhancement was a concept that should be included. That is, if evidence of drug-induced up-regulation of immune function (such as signs of autoimmunity) were observed (especially in clinical trials), this effect should be examined where possible and with available tools. Thus, the document could be interpreted to support the use of such assays as the popliteal lymph node assay (PLNA) where needed. However, the EWG agreed that it was not possible to make specific recommendations concerning testing strategies for drug allergy. The issue of potential protein drug immunogenicity has already been addressed in a previously published ICH document (S6).

Finally, there are a few points that need to be emphasized. The S8 document contains a list of factors that should be considered in a weight-of-evidence approach to deciding when to conduct immunotoxicity testing. Drug class (e.g., anti-inflammatory compounds), effects observed in RTS, observed clinical effects, and intended patient population are among the considerations. However, one issue should be addressed specifically: the role of pathology in looking for signals of immunotoxicity. This has been the source of considerable controversy. Without going into a long discussion about this, suffice it to say that histopathology alone was never intended to be the sole, or even principle, factor in making the decision concerning need for studies. The appropriate methods for “enhance histopathology” are in fact not different from routine practice by veterinary pathologists. The S8 document contains an Appendix with a discussion of various techniques that can be used, but particular attention was given to the issue of pathology and the EWG received excellent advice from several sources. Finally, it should be emphasized that stress is not likely to be considered a sufficient cause for observed adverse effects on the immune system in the absence of compelling evidence. Shrinking thymus in rodents is not compelling evidence. The document should be taken to indicate that immunotoxicity is an important issue that should not be ignored. In fact, it is this issue that, in hindsight, was probably the most important in driving the guidance process for immunotoxicity. There were simply too many examples the EWG knew of where reasons for concern about potential immunotoxicity were apparently ignored. It is anticipated that, with publication of the S8 guidance, this will no longer be the case.

The opinions in this article are those of the author and do not reflect official FDA policy or positions.