By Steve Kennedy* and Anthony D. Sabatelli**
As reviewed previously, antibody-drug conjugates (ADCs) are a highly promising class of antitumor drugs that represent a growing proportion of cancer treatments in the development pipeline (see "Antibody Drug Conjugates: The Patent Landscape for a New Class of Cancer Treatment"). By combining the specificity of monoclonal antibodies targeted to tumor markers with the strong antitumor potential of potent small-molecule drug payloads, ADCs have the potential to aggressively treat many cancers with fewer side effects than traditional therapies. A key component in their design is the linker that joins the antibody and the payload.
The choice of conjugation strategy for attaching the linker and payload to the antibody can have a profound impact on potency, safety, and reproducibility. In another previous article, we reviewed some patents relating to the method of attachment of the drug to the antibody (see "Linking It Up: Antibody-Drug Conjugates"). The other major considerations when choosing or designing a linker are its stability in circulation and ease of payload release at the target. The linker must be resistant to cleavage as the conjugate circulates in the blood to prevent early release of the payload, which could reduce antitumor activity and specificity and also increase side effects. However, the conjugate must also be capable of releasing the payload at the tumor site to maintain antitumor activity. With regard to these considerations, there are two major classes of linkers: noncleavable and cleavable.
Noncleavable linkers are composed of highly stable bonds, providing integrity to the conjugate while in circulation. Noncleavable linkers require the proteolytic degradation of the antibody rather than the linker in order to release the drug payload, typically in the lysosome of the target tumor cells. Because the payload remains attached to an amino acid residue of the antibody, as well as to the linker itself, after antibody proteolysis, the payload must still maintain its effectiveness with these additional moieties still attached. Although a number of payloads have been designed to maintain their potency in this form, notably for the FDA-approved Kadcyla (US8088387B2), noncleavable linkers can present increased research investment relative to cleavable linkers. Kadcyla, is an antibody-drug conjugate consisting of the humanized monoclonal antibody trastuzumab covalently linked to the cytotoxic agent DM1 (mertansine). Generally, ADCs with noncleavable linkers require more extensive pharmacokinetic/pharmacodynamic (PK/PD) modeling to account for the presence of multiple metabolites.
Several noncleavable linkers have been developed, and their effectiveness has been demonstrated in a number of applications. Kadcyla employs a N-succinimidyl-4-(N-maleimidomethyl)cyclohexane-1-carboxylate (SMCC) linker that has also been used for other DM1 conjugates (US8142784B2, WO2013075048A1). The SMCC linker has also been used effectively for conjugating doxorubicin payloads (US6630579B2). Maleimidocaproyl linkers have also been used to conjugate auristatin derivatives to monoclonal antibodies, improving potency without significant PK/PD differences (US8512707B2, US9120854B2). However, SMCC and maleimidocaproyl linkers have solubility limitations. To improve solubility, polyethyleneglycol (PEG) based linkers have also been designed, most notably the sulfo-SPDB linker, which improves both solubility and effectiveness against certain multi-drug resistant tumors (US8088387B2, US9023351B2, US20160095938A1, US9446146B2, WO2005009369A2).
More common than noncleavable linkers are linkers that are cleavable in specific environments or by individual enzymes. Although these linkers are not typically as proteolytically stable as noncleavable linkers, they allow for the conjugation of many varieties of drug payloads without impacting their potency or requiring such extensive PK/PD modeling. One of the earliest cleavable linkers employed, in the first FDA-approved ADC Mylotarg (US5773001), was an acid-sensitive hydrazone that was relatively stable at neutral pH. When this ADC is transported to acidic environments, such as the lysosome, the hydrazone is readily hydrolyzed and releases the payload. This linker has been used both alone and in combination with linkers reducible by glutathione (US6214345B1, US10111954B2, US8153768B2), but they tend to exhibit non-specific drug release in clinical studies. It is believed that this effect is induced by slow linker hydrolysis at physiological conditions, and efforts have been made to develop more-stable cleavable linkers.
One of the most common cleavable linker strategies is to insert dipeptide sequences recognized by the cathepsin B protease. Because this protease is often overexpressed in tumor cells, the linker is readily cleaved and releases its payload once it is internalized at its target. The most common dipeptides used are valine-alanine and valine-citrulline. The valine-citrulline didpeptide is particularly effective since it has improved stability over more commonly-occurring peptide sequences. The FDA-approved Adcetris (US7829531) is a prototypical example, but other ADCs have showed promising preclinical and clinical results using this linker (US20160082119A1). To improve reproducibility between animal models and human trials, the valine-citrulline linker has been further modified in some cases to a glutamic acid-valine-citrulline linker as well (WO2018218004A1). In addition to cathepsin B, a number of ADCs have employed proteolysis by β-glucoronidase, which is present at high levels in lysosomes and some tumors, but otherwise have a similar mode of action (US8568728B2, WO2015057699A2).
Other promising linkers have showed even greater stability than previous cleavable linkers without sacrificing delivery of the drug payload. Although the specific mechanism of action is still being determined, pyrophosphate diester linkages have been used to generate ADCs that are stable in plasma for up to seven days while still rapidly releasing their payloads in the endosome (US20170182181A1). Additionally, thiazolidine linkers originally designed for improved conjugation performance have also shown improved stability and have been used as cleavable linkers (US9198979B2).
As evidenced by the flurry of intellectual property activity surrounding ADC linker chemistries and the growing proportion of ADCs in the clinical pipeline, it is likely that innovations in linker stability and selectivity will continue to provide novel opportunities for the biopharmaceutical industry. Considering the intense research efforts and potential for overlap for ADCs, novel developments and adept patent claiming can be fruitful areas of investment for organizations seeking to advance the field of linker chemistry.
Steve Kennedy is a Ph.D. Candidate in the Chemistry Department at New York University. He specializes in biophysical characterization of protein complexes and is currently focused on the role of adaptor proteins in signaling pathways. Prior to attending NYU, Steve obtained his B.S. in Chemistry with Cum Laude honors at the University of Massachusetts - Boston, during which time he conducted bioanalytical mass spectrometry method development and lipidomics research.
** Dr. Sabatelli is Patent Counsel with Wiggin and Dana LLP
Comments