Clinical proof of concept for a safe and effective NF‐κB‐targeting strategy in multiple myeloma

Abstract

Despite the recent introduction of new treatments and improvements in clinical outcomes, multiple myeloma (MM) remains a medical problem. The majority of patients relapse and die from their disease, underscoring the need for new treatments (Kumar et al, 2017). The nuclear factor (NF)-jB pathway is aberrantly activated in virtually all cases of MM, where it enables tumour cell survival, and is therefore considered an effective therapeutic route in MM. Yet, following an aggressive 30-year effort by the pharmaceutical industry, no specific NF-jB inhibitor has been clinically approved, due to the dose-limiting toxicities associated with the general NFjB suppression. Agents indicated in MM, e.g., proteasome inhibitors and immunomodulatory drugs, inhibit NF-jB, but also many other pathways, and neither specifically target cancer cells nor afford their clinical benefit through NF-jB inhibition (Greten et al, 2007; Di Donato et al, 2012; Begalli et al, 2017; Bennett et al, 2018). Therefore, there is a need for a fresh approach to safely block NF-jB signalling in MM. To overcome the barrier to safe NF-jB inhibition, we developed a novel class of NF-jB-targeting therapeutics. We identified the interaction between the JNK-activating kinase, MKK7, and the NF-jB-regulated antiapoptotic factor, GADD45b (De Smaele et al, 2001; Papa et al, 2004; Tornatore et al, 2014; Capece et al, 2018), as an essential, cancerrestricted survival module downstream of NF-jB in MM. Accordingly, GADD45b is selectively highly expressed in MM cells, where it denotes shorter patient overall survival (Tornatore et al, 2014). We developed the first-in-class GADD45b/ MKK7 inhibitor, DTP3, which selectively kills MM cells by inducing MKK7/JNK-dependent apoptosis, ex vivo and in vivo, without toxicity to normal tissues (Tornatore et al, 2014). Further preclinical investigations demonstrated that DTP3 combines on-target-selective pharmacology and favourable drug-like properties with tolerability and none of the preclusive toxicities of conventional IKKb/NF-jB-targeting agents (Tornatore et al unpublished observations). Due to this unique mode of action, DTP3 represents a significant opportunity for managing MM patients. The success rate for novel drug candidates in oncology from first-in-human trial to market registration is only ~5%, due to the heterogeneity of tumours and absence of drugspecific biomarkers, clearly demonstrating therapeutic target engagement (Kola & Landis, 2004). Considering that any GADD45b/MKK7-targeting drug would potentially benefit only a discrete subset of patients, we conducted a preclinical proof-of-concept study (REC 11/LO/1628) to develop a companion biomarker platform capable of assessing pharmacodynamic response in the earliest stages of the DTP3 clinical development. The current Letter reports the clinical proof-of-concept for an NF-jB-targeting strategy as a safe and mechanistically effective novel therapy in MM patients. We sought to first verify the cancer-selective mechanistic specificity of DTP3 in primary MM cells (Supplementary Methods). DTP3 was effective in inducing JNK phosphorylation, denoting therapeutic target engagement and caspase-3 cleavage, an apoptosis hallmark, in malignant CD138 cells, ex vivo (Fig 1A). By contrast, DTP3 produced no such effects in healthy CD20 cells nor in peripheral blood mononuclear cells (PBMCs) from the same patients, even at 30-fold higher concentrations (Fig 1B, C). Given the wide distribution and overall high levels of GADD45B expression in MM cells (Fig 1D, Supplementary Figure S1), we evaluated whether the cancer-selective pharmacodynamic response to DTP3 depended on GADD45B expression. As shown in Fig 1E, in 13 samples of malignant CD138 cells which responded to DTP3, GADD45B expression was significantly higher than in any of the unresponsive CD138-cell samples or any sample of normal CD20 cells or PBMCs. Hence, the capacity of DTP3 to trigger JNK-driven apoptosis in primary MM cells correlates with a high significance with the GADD45B expression levels (Fig 1E). These results demonstrate that the on-target-specific therapeutic response to DTP3 can be monitored and potentially predicted by an objective measurement of GADD45B expression and JNK-dependent apoptosis in tumour cells. Interestingly, upon clinical validation, this approach could inform a patient stratification platform predictive of objective clinical response to support the clinical development of DTP3. Accordingly, we initiated the first-in-human phase-I/IIa trial of DTP3 to evaluate the safety, tolerability, pharmacokinetics and pharmacodynamics of this novel NF-jB-targeting therapeutic in patients with relapsed or refractory MM (EudraCT: 2015-003459-23; Supplementary Methods). Three single-patient, dose-escalation cohorts have been investigated to date, evaluating DTP3 at the dose levels of 0 5, 1 and 2 mg/kg, administered by rapid intravenous infusion three times a week (Fig 2A, Supplementary Figure S2). All patients had progressive disease at the point of entry into the study, despite having received multiple prior lines of therapy. All three patients completed their first 28-day treatment cycle with no complications. No reportable adverse effects, nor