October 1, 2018 (Vol. 38, No. 17)
Julia Stevens Ph.D. Senior Scientist PHASEZERO® Research Services, at BioIVT
Amanda J. Woodrooffe Ph.D. Vice President and General Manager PHASEZERO® Research Services, at BioIVT
BioIVT Shows How Preclinical Studies Can Support IND Applications and CTAs
It is extremely costly to have a therapeutic antibody or antibody-like molecule fail during clinical development. Therefore, drug developers need to mitigate any potential issues before first-in-human studies begin. Many preclinical investigations that are essential components of a regulatory Investigational New Drug (IND) application or Clinical Trial Application (CTA) are intended to minimize the risk of harm when a drug is first given to humans. One such preclinical safety study is the assessment of tissue cross-reactivity (TCR).
A TCR study incorporates a series of ex vivo immunohistochemical (IHC) screening assays conducted primarily to identify off-target binding, but also to pick up previously unknown sites of on-target binding for a novel biotherapeutic. Essentially, the presence of therapeutic antibody binding in frozen ex vivo tissues is used to give an indication of potential organ toxicity in vivo.
Whether such staining actually correlates with organ toxicity is still regularly debated, but TCR studies remain a requirement in the data package for an IND/CTA submission for most biotherapeutics. TCR studies are also used to compare staining patterns between human and animal tissue, providing additional justification for the choice of models used to generate other preclinical safety data.
One of the most important aspects of a TCR study is the development of the IHC protocol. Given that the biotherapeutic to be tested has not been designed as an IHC tool, this poses a technical challenge that can be overcome only by rigorous assay workup. Since a favorable dataset for a TCR study is a negative result, it is very important to know that the assay is specific and robust before the test tissues are examined.
Many researchers underestimate the length of time needed to develop a sound IHC method. Time-pressed researchers may be tempted to cut corners, but they should know better than to risk gathering misleading TCR results when the good laboratory practice (GLP) assays are run.
Test Items come in many forms, some differing substantially from immunoglobulins in structure. Irrespective of format, consideration must be given to how the molecule is to be detected in an IHC assay.
Unlabeled human or humanized antibodies can be detected by precomplexing with an anti–human antibody before applying to the test tissues. However, it is technically easier to work with an antibody that has been labeled with a small molecule, such as biotin or fluorescein isothiocyanate (FITC), that can be detected with an antibody specific for the label. Indeed, for many molecules, labeling is a necessity to facilitate detection.
While biotinylation is a well-established and relatively straightforward technique for labeling biological molecules, FITC may be preferable in a TCR study as it avoids the need for additional avidin-biotin blocking steps in the assay protocol, without which background staining of tissue is likely to prove problematic. In either case, it is important to establish the impact of any labeling on the binding properties of the molecule and to ensure test and control items are labeled to the same degree.
The inclusion of a Control Item is strongly recommended.1 The Control Item, essentially an isotype control, is usually a molecule identical in structure to the test item (including labels such as biotin or FITC) but raised against a molecule unlikely to be found in human tissue—for example, green fluorescent protein, a plant protein, or even snake venom. Ideally, such a molecule should be prepared in parallel with the Test Item and used to ascertain the background level and pattern of tissue binding that occurs irrespective of the complementarity-determining region (CDR).
Failure to appreciate the value of including a suitable Control Item for comparison may result in any binding of the Test Item being mistakenly interpreted as specific and therefore indicative of potential organ toxicity. There are other ways that binding specificity can be assessed, for example, competing out binding by preincubation with a molar excess of soluble antigen. However, these methods add to the overall study cost, and sufficient soluble antigen is not always available.
Positive Control Material
Selection of suitable positive control material is crucially important for TCR protocol development, as well as for use in GLP studies, to validate the test item in all of the assay runs—which are likely to be numerous.2 The positive control material needs to be representative of the frozen tissue that the test item will encounter in the full GLP assays.
Using overexpressing cell lines or other types of positive control material is inferior to using frozen tissue, which retains tissue matrix. Where no suitable tissue exists that naturally expresses the target of interest, various alternative techniques have been attempted, including subcutaneous injection of antigen-coated beads prior to harvesting skin from mice.
To address this issue, we developed a proprietary method by drawing on our expertise in handling human tissue. Our method allows incorporation of soluble antigen into a human tissue matrix, resulting in a frozen, sectionable material suitable for use as a positive control material when no other solution exists.
The quality of the frozen tissues used in GLP TCR studies is of the utmost importance. Whether from postmortem or surgical origin, human tissues must have good morphological preservation to allow adequate interpretation of staining patterns and crucially must retain antigenicity. Confirmation of tissue antigenicity as part of the GLP study is strongly recommended, as testing prior to the GLP study does not guarantee antigenicity as a sample is sectioned through (Figures 1A–1C).
Interpretation of Results
A qualified pathologist should interpret the results. Staining observed with the Test Item should be compared to that seen in adjacent sections incubated with the Control Item, and specific staining should be considered only where Control Item staining is not present. The pathologist may, in addition, make a judgment as to whether smeared or very diffuse staining is specific in nature.
The cellular location of specific staining is also important to consider, since staining of cytoplasm, which is unlikely to be accessible to a biotherapeutic in vivo, is less likely to translate into a biological effect than membrane staining, so may be less of a concern from a safety perspective.
The biological relevance of any TCR staining can never be fully determined until other human safety/toxicity data (for example, from clinical trials or post-marketing surveillance) become available, so it is important to interpret data with caution. TCR studies are here to stay, at least for now. It will be interesting to see what regulatory bodies, such as the FDA and EMA, will require for tissue cross-reactivity assessment as newer technologies become available.
1. Leach MW, Halpern WG, Johnson CW, Rojko JL, MacLachlan, TK, Chan CM, Galbreath EJ, Ndifor AM, Blanset DL, Polack E, Cavagnaro JA. Use of tissue cross-reactivity studies in the development of antibody-based biopharmaceuticals: History, experience, methodology, and future directions. Toxicol. Pathol. 2010; 38: 1138–66.
2. Geoly FJ. Tissue Cross-Reactivity Studies: What Constitutes an Adequate Positive Control and How Do We Report Positive Staining? [Regulatory Forum Opinion Piece.] Toxicol. Pathol. 2014; 42: 954–956.