c-KIT overexpression is well recognized in cancers such as gastrointestinal stromal tumors (GIST), small cell lung cancer (SCLC), melanoma, non-small cell lung cancer (NSCLC), and acute myelogenous leukemia (AML). Treatment with the small-molecule inhibitors imatinib, sunitinib, and regorafenib resulted in resistance (c-KIT mutant tumors) or limited activity (c-KIT wild-type tumors). We selected an anti-c-KIT ADC approach to evaluate the anticancer activity in multiple disease models. A humanized anti-c-KIT antibody LMJ729 was conjugated to the microtubule destabilizing maytansinoid, DM1, via a noncleavable linker (SMCC). The activity of the resulting ADC, LOP628, was evaluated against GIST, SCLC, and AML models and against GIST and SCLC models. LOP628 exhibited potent antiproliferative activity on c-KIT-positive cell lines, whereas LMJ729 displayed little to no effect. At exposures predicted to be clinically achievable, LOP628 demonstrated single administration regressions or stasis in GIST and SCLC xenograft models in mice. LOP628 also displayed superior efficacy in an imatinib-resistant GIST model. Further, LOP628 was well tolerated in monkeys with an adequate therapeutic index several fold above efficacious exposures. Safety findings were consistent with the pharmacodynamic effect of neutropenia due to c-KIT-directed targeting. Additional toxicities were considered off-target and were consistent with DM1, such as effects in the liver and hematopoietic/lymphatic system. The preclinical findings suggest that the c-KIT-directed ADC may be a promising therapeutic for the treatment of mutant and wild-type c-KIT-positive cancers and supported the clinical evaluation of LOP628 in GIST, AML, and SCLC patients. .

Embryonic (before the end of organogenesis) exposure was assessed in one molecule each in rabbit, rat, and mouse, but detectable levels were present only in rodents. In rodents, fetal levels remained relatively constant from gestation day (GD) 16 and 17 until the end of gestation, while maternal levels decreased or remained constant in rat and decreased in mice. In rabbits, following a last dose on GD 19, fetal levels increased markedly in late gestation while maternal levels decreased. In the cynomolgus monkey, fetal levels increased substantially from GD 50 to 100 and were maintained relatively constant through parturition (approximately GD 165). Based on available data of both the monkey and rabbit, IgG1 molecules appeared to transfer more readily than other isotypes in late gestation. Across all species, there was no differential transfer based on pharmacologic target being soluble or membrane bound. Within each species there was a correlation between maternal and fetal exposure, suggesting it may be possible to predict fetal exposures from maternal exposure data. These nonclinical data (both temporal and quantitative aspects) are discussed in a comparative context relative to our understanding of IgG placental transfer in humans.

Hypersensitivity reactions (HSRs) were observed in three patients dosed in a phase I clinical trial treated with LOP628, a KIT targeted antibody drug conjugate. Mast cell degranulation was implicated as the root cause for the HSR. Underlying mechanism of this reported HSR was investigated with an aim to identifying potential mitigation strategies. Biomarkers for mast cell degranulation were evaluated in patient samples and in human peripheral blood cell-derived mast cell (PBC-MC) cultures treated with LOP628. Mitigation strategies interrogated include pretreatment of mast cells with small molecule inhibitors that target KIT or signaling pathways downstream of FcεR1, FcγR, and treatment with Fc silencing antibody formats. Transient elevation of serum tryptase was observed in patients 1-hour posttreatment of LOP628. In agreement with the clinical observation, LOP628 and its parental antibody LMJ729 induced degranulation of human PBC-MCs. Unexpectedly, KIT small molecule inhibitors did not abrogate mast cell degranulation. By contrast, small molecule inhibitors that targeted pathways downstream of Fc receptors blunted degranulation. Furthermore, interference of the KIT antibody to engage Fc receptors by pre-incubation with IgG or using engineered Fc silencing mutations reduced or prevented degranulation. Characterization of Fcγ receptors revealed human PBC-MCs expressed both FcγRII and low levels of FcγRI. Interestingly, increasing the level of FcγRI upon addition of IFNγ, significantly enhanced LOP628-mediated mast cell degranulation. Our data suggest LOP628-mediated mast cell degranulation is the likely cause of HSR observed in the clinic due to co-engagement of the FcγR and KIT, resulting in mast cell activation. .

Non-cleavable linkers are used in a number of different modalities for various reasons, such as linking an active drug moiety to half-life extending molecules, to groups that enable a specific tissue or cell targeting or to facilitate active uptake into target cells. Non-cleavable linkers do not have a designated weak point in their structure that can lead to cleavage by proteases, hydrolases or chemically by pH changes. Consequently, when designing a conjugate, the choice of a non-cleavable over a cleavable linker is usually a consequence of pursuing a certain mode of action where the stability of the complex is more important than a fast liberation of the active moiety. Linkers of various length, polarity, stability and flexibility are used for different types of conjugates and the linker design is mostly driven by the particular purpose and desired mode of action. This article reviews non-cleavable linkers applied predominantly in Antibody Drug Conjugates (ADCs), and how they influence these conjugates in terms of ADME properties (absorption, distribution, metabolism and elimination) and safety.

Authors presently affiliated with Novartis are full-time employees of Novartis. Authors not presently affiliated with Novartis have their current affiliations listed. Where the current affiliation is a private company the authors are full-time employees of this company.