HDP-101, an Anti-BCMA Antibody–Drug Conjugate, Safely Delivers Amanitin to Induce Cell Death in Proliferating and Resting Multiple Myeloma Cells

Multiple myeloma is a malignancy characterized by monoclonal plasma cell (PC) expansion in the bone marrow (1, 2). Despite recent improvements in progression-free survival (PFS) and overall survival (OS) following the introduction of new agents (3), most patients with multiple myeloma eventually relapse.

Diverse immunotherapeutic approaches including monoclonal antibodies, T-cell–bispecific antibodies, CAR-T cell therapies, and antibody–drug conjugates (ADC) are promising new assets in multiple myeloma therapy (4, 5). Monoclonal antibodies have proven efficacy in clinical trials for treating multiple myeloma, with the best results being achieved in combination with standard-of-care drugs (6, 7). Bispecific antibodies have shown promising early results (8). Nevertheless, they carry the risk of adverse immune reactions and depend on intact T-cell function which is impaired in patients with refractory multiple myeloma (9, 10). Initial data on CAR-T cell therapies are highly promising; however, patient numbers are small and response durations are shorter than expected. Furthermore, CAR-T cell therapies require costly individual ex vivo manufacturing for each patient (11, 12). For these reasons, ADCs may represent an alternative to multiple myeloma immunotherapy (13).

B-cell maturation antigen (BCMA), a cell surface receptor recognizing B-cell activating factor (BAFF) and a proliferation-inducing ligand (APRIL), play an important role in the survival of long-lived PCs in the bone marrow (14–17). BCMA is expressed on PCs and is highly expressed on myeloma cells while being absent on most other cells (18). Patient serum levels of soluble BCMA (sBCMA) correlate with disease status, response to therapy, and OS (19, 20). Due to this favorable pattern of expression, BCMA is highly suitable for targeted therapy in multiple myeloma, such as using ADCs.

Alpha-amanitin, a bicyclic peptide of eight amino acids, is a eukaryotic RNA polymerase II inhibitor, arresting the cellular transcription process and protein synthesis even at very low concentrations (21–23). This inevitably leads to cell death by apoptosis (24). Due to its hepatotoxicity (25) and renal toxicity (26), it has not to date undergone clinical evaluation. Recently, when used as “payload” of an ADC, it demonstrated promising preclinical efficacy with limited toxicity (27, 28). Most available ADCs are based on toxins with other mechanisms of action: microtubuli- (auristatins and maytansines) or DNA-targeting toxins (duocarmycins and pyrrolobenzodiazepines). Spindle poisons require active cell proliferation to exert their therapeutic benefit and are ineffective in resting cells. In contrast, alpha-amanitin deploys its cytotoxicity irrespective of the proliferation status of tumor cells due to its inhibitory effect on DNA transcription.

In pursuit of new treatment strategies, the compound HDP-101 was developed; this is an ADC targeting BCMA with an amanitin derivative as the conjugated toxin. Here, we report on the cytotoxic potency of HDP-101, both in vitro and when using in vivo models of multiple myeloma, and including primary cells from patients with multiple myeloma refractory to proteasome inhibitors, immunomodulatory drugs, and monoclonal antibodies. Furthermore, we show HDP-101 is active against both proliferating and resting multiple myeloma cells while being well-tolerated by cynomolgus monkeys in a tolerability study. Amanitin-based ADCs, also called antibody-targeted amanitin conjugates (ATAC), are a promising novel approach in the therapy of multiple myeloma with a distinct mode of action to overcome drug resistance and to improve patient outcomes.

Sequence of the anti-BCMA antibody (J22.9-ISY) and the chemical structure of the toxin linker are described in WO2018115466A1 (SEQ ID NO: 1 = heavy chain; SEQ ID NO: 2 = light chain; Formula 1 = toxin linker). The cysteine reactive linker-amanitin compound HDP 30.2115 with the cleavable linker was conjugated to engineered cysteine residues of anti-BCMA and anti-DIG (digoxigenin) THIOMAB antibodies using maleimide chemistry by Heidelberg Pharma Research GmbH (HDPR). In brief, THIOMABs in PBS, pH 7.4, were reduced with TCEP, and interchains disulfides were reoxidized by dehydroascorbic acid. Subsequently, the engineered cysteines were used for conjugation with cysteine reactive linker-amanitin compound HDP 30.2115. The drug–antibody ratio (DAR) according to LC-MS analysis was 2 toxins per IgG. Monomethyl auristatin F (MMAF) and phycoerythrin (PE) conjugated to interchains of anti-BCMA antibodies with noncleavable linkers were also generated using maleimide chemistry by HDPR. In brief, antibody in PBS, pH 7.4, was reduced with TCEP and conjugated to MMAF and PE. The conjugates were purified by SE-FPLC and dialysis.

HDP-101–binding affinity to human BCMA was analyzed using the FortéBIO Octet K2 system (FortéBio). Anti-hIgG Fc Capture Biosensors were used for HDP-101 binding at different concentrations as the ligand, human BCMA served as analyte. The dissociation constant (K_D) value was calculated using the FortéBIO Data Analysis 10.0 software.

Cells were seeded and drugged at a density of 2 × 10⁴ cells/mL in 96-well plates and incubated for 48 or 96 hours. Cytotoxicity of HDP-101, MMAF-coupled anti-BCMA antibody, anti–DIG ATAC, unloaded BCMA antibody, alpha-amanitin, carfilzomib, dexamethasone, doxorubicin, melphalan, and pomalidomide was determined using the cell viability assay CellTiter-Glo 2.0 (Promega) according to the manufacturer's instructions. Then compounds' IC₅₀ (drug half-maximal inhibitory concentration at which 50% of its maximum effect is observed) was calculated. For spike-in experiments with sBCMA, NCI-H929 and KMS-11 cells were treated with HDP-101 in the presence or absence of 100 or 1000 ng/mL sBCMA, produced by HDPR, and incubated for 96 hours. Cytotoxicity was measured using the cell viability assay CellTiter-Glo 2.0 (Promega). Carfilzomib, dexamethasone, doxorubicin, melphalan, and pomalidomide were purchased from Selleck Chemicals, and free alpha-amanitin was purchased from AppliChem. Data analysis was performed with GraphPad Prism 7 (GraphPad Software, Inc.) software to plot curve fits and perform statistical analyses. Primary multiple myeloma cells were obtained after written-informed consent from patients with approval of the institutional ethics committee according to the Declaration of Helsinki (further Information on used cell lines can be found in the Supplementary Materials and Methods).

For internalization assays, anti-BCMA antibody and HDP-101 were incubated with MM.1S, MM.1S Luc, L363, OPM-2, NCI-H929, U266B1, and RPMI-8226 cells at 4°C (total cell surface binding) and 37°C (internalization) for 2 hours. The cellular fluorescence with and without the Alexa 488–quenching antibody (Invitrogen) was determined by FACS using goat anti-Human IgG (H+L)-Alexa Fluor 488 F (ab') 2-Fragment (Dianova) as a secondary antibody on a FACSCalibur (BD Biosciences). Internalization was calculated as quenched samples fluorescence divided by unquenched samples fluorescence (corrected for untreated cells).

Medium supernatants were collected from multiple myeloma cell line cultures (U266, RPMI-8226, OPM-2, MM.1S, MM.1S Luc, and NCI-H929); patient BMSCs, BMSC-U266 and BMSC-MM.1S, coculture 4 days after cell splitting and diluted 1:100. Human BCMA levels were determined using an ELISA kit (R&D Systems).

HDP-101 stability was tested following incubation at 37°C for 0, 4, and 10 days in human, mouse, or cynomolgus plasma. The analysis was performed by amanitin Western blot and cytotoxicity assays on NCI-H929 cells.

Two groups of 10 mice each were used in each experiment. Six- to 8-week-old female SCID beige mice were obtained from Charles River. NCI-H929 subcutaneous: Animals were implanted with 5 × 10⁶ NCI-H929 cells into the flank. When tumors reached a size of 100 to 200 mm³, animals received a single i.v. dose. Tumor volumes were monitored twice per week and were calculated as V = a × b² × 0.5, where a is the length and b is the width. MM.1S Luc i.v.: Animals were implanted with 1 × 10⁷ MM.1S Luc cells i.v. into their tail vein. Once a mean total flux of 1.5 × 10⁶ to 1 × 10⁷ (∼14 days after implantation) was reached, animals received a single i.v. dose. Luciferase activity was monitored by noninvasive bioimaging (Caliper IVIS).

Nonhuman primates (NHP) studies were performed in treatment-naïve cynomolgus monkeys at LPT. A 3-cynomolgus-monkey group (age 2–5 years, 2.6–3.1 kg bodyweight at 1st dosing) was kept according to German animal protection law requirements, and all animal experiments were approved by the relevant competent authorities. Female animals were chosen for this study, and there was no prejudice to exclude males. Animals received HDP-101 i.v. (slow bolus) in escalating doses (0.3, 1, and 3 mg/kg; doses), and serum was obtained at different time points. Animals were monitored over time for biochemical and hematologic blood parameters, urinalysis, body weight, food consumption, clinical signs, and mortality. At the study end, animals were sacrificed and subjected to histopathologic evaluation.

The data are shown as mean ± SD or SEM. In vitro experiments were done in triplicates and repeated at least twice. Statistical significance of differences observed was determined using the Student t test. Survival data were plotted by the Kaplan–Meier method and analyzed by the log-rank test using Prism Version 7 (GraphPad Software). Minimal significance considered was P < 0.05.

Additional Supplementary Materials and Methods are available.

Target epitope selective binding is essential for the ADC therapeutic efficacy. Therefore, we analyzed the HDP-101 to BCMA-binding capability by FACS analysis on BCMA-expressing cell lines. NCI-H929, RPMI-8226, and MM.1S cells were incubated with unconjugated anti-BCMA THIOMAB control antibody or HDP-101 increasing concentrations. The antibody like HDP-101 showed specific binding to the BCMA⁺ cells (Fig. 1A). Affinity measurement revealed anti–BCMA antibody-binding likewise of HDP-101 to human BCMA with a mean K_D of 2.32 and 2.66 nmol/L, respectively (Fig. 1B).

Furthermore, upon binding to the target effective ADC-target complex internalization into the tumor cell is required to release amanitin to exert its cytotoxic effects in multiple myeloma cells. The HDP-101 internalization and the anti-BCMA antibody were investigated by measuring cellular fluorescence by FACS analysis in seven BCMA-positive cell lines. Both, the anti-BCMA THIOMAB and HDP-101 were rapidly internalized after 2 hours to a similar extent (Fig. 1C).

Taken together, HDP-101 efficiently binds to human BCMA and is internalized by BCMA-positive multiple myeloma cell lines.

The humanized anti-BCMA antibody was engineered using THIOMAB technology to facilitate site-specific conjugation with reactive cysteine residues for the production of an ADC with a controlled DAR. An amino acid exchange to substitute aspartic acid in the heavy-chain region at position 265 (D265C) was chosen for cysteine substitution. This results in site-specific conjugation of 2 toxin molecules per monoclonal antibody (Supplementary Fig. S1A). Furthermore, although the unmodified anti-BCMA antibody showed antibody-dependent cellular cytotoxicity (ADCC) activity to a certain extent, no measurable anti-BCMA ADCC activity was detected using HDP-101 in an ADCC reporter bioassay (Supplementary Fig. S1B). These results confirm the anti-BCMA THIOMAB, and in turn, HDP-101 causes only minimal, if any, antigen-independent cytotoxicity.

We examined the HDP-101 cytotoxic effect in vitro on BCMA⁺ multiple myeloma cell lines (U266, RPMI-8226, MM.1S, NCI-H929, OPM-2, INA-6, SKMM.1, L363, and LP-1) likewise on the BCMA-negative bone marrow stroma cell line HS-5 and the sarcoma cell line U2OS. Overall, HDP-101 showed cytotoxic effects on all multiple myeloma lines with IC₅₀ in the nanomolar and subnanomolar range (Fig. 2A and B). To test the HDP-101 specificity, negative control with an antibody against DIG was conjugated with the same amanitin-linker derivative HDP 30.2115 (anti–DIG ATAC). This control anti–DIG ATAC showed no cytotoxic activity up to a concentration of 10⁻⁶ mol/L for most cell lines with slightly reduced viability of U266 and SKMM.1 cells at 10⁻⁷ mol/L (Fig. 2C).

To confirm the mode of cell death in multiple myeloma cells, activation of apoptosis effector caspase 3 and PARP-1 caspase substrate cleavage was measured (29, 30) after exposure to HDP-101 at concentrations ranging from 0.8 to 135 nmol/L for 48 and 72 hours, respectively. Multiple myeloma cell lines U266, MM.1S, and OPM-2 showed dose-dependent PARP and caspase 3 cleavages (Fig. 2D). In parallel, caspase 3 and 7 activations were detected in a dose-dependent manner using the caspase 3/7 assay, confirming that HDP-101 induces cell death via apoptosis in multiple myeloma cells (Supplementary Fig. S2A).

In contrast, cell-cycle analysis using flow cytometry analysis did not reveal a specific accumulation pattern within any stage cell cycle of multiple myeloma cell lines (U266, MM.1S) incubated with 1 nmol/L HDP-101 for 0, 24, 48, 72, and 96 hours (Supplementary Fig. S2B).

Recent studies suggested that multiple myeloma cells shed BCMA, and sBCMA concentration interferes with the BCMA-targeted therapies efficacy. First, we quantified PC surface BCMA (Fig. 2E) and confirmed target expression by all multiple myeloma cell lines albeit at a wide range (1,171–8,987 of antibody molecules bound per cell). We also tested sBCMA levels in multiple myeloma cell line cultures in correlation to their respective HDP-101 efficacy (Fig. 2F). Measuring sBCMA in the multiple myeloma cell lines supernatant revealed a positive correlation (R² = 0.929) between PC surface BCMA and sBCMA (Fig. 2G). Interestingly, NCI-H929 cells with higher sBCMA concentration (401.4 ng/mL) in the supernatant displayed higher sensitivity toward HDP-101 (IC₅₀, 0.03 nmol/L) in comparison with RPMI-8226, which had the lowest sBCMA concentration (42.81 ng/mL) and showed reduced cytotoxicity (IC₅₀, 141.9 nmol/L), indicating target expression might have a greater impact on efficacy than sBCMA concentration.

Spike-in of sBCMA to cytotoxicity assay of multiple myeloma cell lines NCI-H929 and KMS-11 with HDP-101 revealed sBCMA dose-dependent reduction of HDP-101–induced cytotoxicity (Fig. 2H). However, even with a very high spike-in concentration of sBCMA, HDP-101 was still able to completely eradicate multiple myeloma cell lines within in vivo achievable dose ranges (Fig. 2H).

To assess the plasma stability in vitro, HDP-101 was incubated in human, murine, and cynomolgus plasma for up to 10 days. After the respective incubation period, HDP-101 stability was either analyzed in an anti-amanitin Western blot (Supplementary Fig. S3A) or in a cytotoxicity assay on NCI-H929 cells to determine the remaining cytotoxic potential (Supplementary Fig. S3B). The Western blot analysis showed high stability of HDP-101 in human, mouse, and cynomolgus plasma with slightly decreased signals after 10 days. This translated to an only marginally decreased cytotoxic effect of HDP-101 on NCI-H929 cells after 10 days of plasma incubation.

Antitumor activities of ADCs were determined in an NCI-H929 s.c. xenograft model and an MM.1S Luc i.v. xenograft model in vivo (Fig. 3A). A single treatment with 2.2 mg/kg doses only resulted in an initial response followed by tumor regrowth. At 4.4 mg/kg doses, complete tumor remission was reached in 7 of 8 mice for HDP-101, and mice stayed tumor-free for more than 100 days (i.e., the experiment duration; Fig. 3B). No statistically significant bodyweight reduction was observed, and the survival rate in 2.2 mg/kg group was 75% (day 87) and 100% in the 4.4 mg/kg group.

In the disseminated MM.1S Luc xenograft model, doses of HDP-101 were further reduced to a single-dose application of 0.1, 0.3, 1, and 2 mg/kg (Fig. 3C). Single-dose application of 0.1 or 0.3 mg/kg resulted in tumor response below detection limits up to 60 days. At 1 and 2 mg/kg, 9 to 10 animals remained tumor-free for the experiment duration (92 days; Fig. 3D and E). The survival rate for both groups was 100% on day 91, and no statistical bodyweight reduction was observed.

The HDP-101 MTD was investigated in CB17 SCID mice (n = 3) after single i.v. administration of doses up to 40 mg/kg. The HDP-101 MTD was determined as 30 mg/kg which corresponds to 538 μg/kg bound amanitin.

The HDP-101 tolerability was further assessed in a non-GLP dose-escalating tolerability study in the cynomolgus monkey. One group consisting of 3 female animals was injected with HDP-101 in an increasing and repeated dose setting using 0.3 mg/kg (day 1), 1 mg/kg (day 21), and 4 doses of 3 mg/kg (days 42, 64, 84, and 106). Neither signs of systemic intolerance nor any changes in body weight at all dose levels were detected. A set of biochemistry parameters, including total bilirubin, aspartate aminotransferase, alanine aminotransferase (ALT), alkaline phosphatase, total protein, albumin/globulin, and gamma-GT, were examined. The most sensitive parameters were ALT and lactate dehydrogenase (LDH; Fig. 4A), and other parameters remained within the normal range. The highest nonseverely toxic dose was found to be 3 mg/kg in a 3-week cycle of repeated dosing.

Concentrations of HDP-101 and free toxin were determined in cynomolgus serum samples at different time intervals ranging from 5 minutes up to 21 days after each injection. The HDP-101 serum concentrations and free toxin showed a dose-dependent correlation with injected doses of 0.3, 1, and 3 mg/kg. Besides, maximum serum concentrations after each repeated dose injection (3 × 3 mg/kg; RD1, RD2, and RD3) were in a similar range, and no accumulation was observed. Between the repeated applications of 3 mg/kg, HDP-101 serum concentration was 20 μg/mL (Fig. 4B). Serum concentrations of released toxin were found only in the range of lower limit of quantification and were highly variable, accordingly (Fig. 4B).

The histopathologic evaluation at the test period end showed only the following signs of toxicity: two animals presented a mild to marked interstitial nephritis with hyaline tubular casts, cortical fibrosis, and necrosis of tubular cells. One of these animals additionally showed a thickening of Bowman's capsule in the kidneys glomeruli. There were no findings in the liver. All 3 monkeys survived the test period of 126 days.

With regard to liver and nephrotoxicity, another study in cynomolgus monkeys was performed. In this study, doses of 1.0, 2.0, and 3.0 mg/kg of HDP-101 were applied either as single dose or repetitive dosing (4 times, every 21 days) to elucidate differences in liver and kidney toxicity on single versus multiple dosing. Treatment with HDP-101 produced dose-dependent pathology and/or clinical pathology findings in the kidney. However, findings of hyaline tubular casts, interstitial cortical fibrosis, and hyaline eosinophilic droplets in the kidney showed complete or partial recovery. Although there were no correlating histopathology findings in the liver at any dose level, transient elevations of liver biomarkers are considered adverse at multiple doses of ≥2 mg/kg HDP-101 due to the severity of the increase. All pathology findings showed partial or complete recovery after a treatment-free period indicating reversibility of these findings.

To assess HDP-101 specificity for primary BCMA-expressing cells, peripheral blood mononuclear cells (PBMC) from healthy donors were treated with increasing concentrations of HDP-101. No cytotoxic effects were observed up to a concentration of 0.1 μmol/L, indicating a favorable therapeutic window (Fig. 5A). Likewise, neither free amanitin nor the unconjugated BCMA antibody in the same dose range showed relevant signs of cytotoxicity (Supplementary Fig. S4A and S4B).

A wide range of cell surface BCMA epitope expression was detected on primary human BMPCs from patients with multiple myeloma irrespectively of disease status (Fig. 5B). Next, the HDP-101 efficacy on primary multiple myeloma cells from patients with newly diagnosed (n = 5) disease (median IC₅₀, 1.1 nmol/L) and relapsed/refractory disease after one to three (n = 6; median IC₅₀, 0.14 nmol/L) or more than three prior treatment therapies (n = 7; median IC₅₀, 0.98 nmol/L) was tested, respectively (Fig. 5C). HDP-101 showed cytotoxic effects on BMPCs, independent of prior lines of treatment (Fig. 5C).

Having shown HDP-101 specifically binds to BCMA, we next quantified BCMA molecules' abundance on the cell surface of multiple myeloma cell lines and patient BMPCs by flow cytometry using PE-labeled BCMA antibodies measuring the antibodies bound per cell (ABC). In general, multiple myeloma cells displayed a wide range of BCMA epitope density (1170.62–8987.25). Multiple myeloma cell lines had significantly (P < 0.01) higher BCMA epitope density (Fig. 2E) compared with cells of patients with both newly diagnosed and relapsed multiple myeloma (Fig. 5B).

Importantly, no correlation between BCMA abundance on the cell surface and sensitivity toward HDP-101 (depicted by IC₅₀) could be shown in primary multiple myeloma cells (correlation index R², 0.1034), indicating HDP-101 requires only very low BCMA abundance on the cell surface for effective cell killing (Supplementary Fig. S4C).

Nondividing tumor cells are difficult to treat as they are less sensitive to cell-cycle–dependent compounds, such as microtubule- or DNA-targeting toxins. In this context, primary multiple myeloma cells typically show a low growth rate in vivo and do not proliferate in vitro. Therefore, we further examined the amanitin effect as the payload compared with the widely used antimitotic agent, monomethyl auristatin F (MMAF), on nondividing multiple myeloma cells. BMPCs from patients with multiple myeloma were exposed to control media, HDP-101, or the identical anti-BCMA antibody but coupled with MMAF at concentrations ranging from 1.28 pmol/L to 0.1 μmol/L (Fig. 6A). HDP-101 but not the anti–BCMA-MMAF ADC induced almost complete BMPCs' cell death. In addition, the multiple myeloma cell line U266 was cultured in starvation medium (RPMI-1640 + 0.1%FCS) to stop cell proliferation. Resting U266 cells were then treated with HDP-101 or the anti–BCMA-MMAF ADC at concentrations ranging from 1.28 pmol/L to 0.1 μmol/L for 96 hours (Fig. 6A). Although starvation only reduced HDP-101 cytotoxicity on U266 cells at the lowest dose (0.16 nmol/L, P < 0.001), it led to significant treatment resistance in U266 cells treated with the anti–BCMA-MMAF ADC (0.16 nmol/L, P < 0.01; 4 nmol/L, P < 0.001; 0.1 μmol/L P < 0.0001; Fig. 6B). Taken together, our results show, in contrast to MMAF, amanitin as a payload has a strong cytotoxic effect on nonproliferating multiple myeloma cells.

The tumor microenvironment is known to confer protective effects on multiple myeloma cells against cytotoxicity by cytokine secretion such as IL6 or IGF-1 likewise through cell adhesion–mediated drug resistance (CAM-DR; refs. 31, 32). Therefore, multiple myeloma cell lines (U266 and MM.1S) were cultured in IL6 or IGF-1 presence or absence for 96 hours while being exposed to increasing concentrations of HDP-101 (1.28 pmol/L to 0.1 μmol/L; Fig. 7A). Although the cytokines stimulated cell proliferation, no significant changes in IC₅₀ values were observed in either cell line compared with those without cytokines. To test for CAM-DR in the context of HDP-101 exposure, the multiple myeloma cell line MM.1S Luc was cocultured in the presence of patient BMSCs with HDP-101 increasing doses for 96 hours (Fig. 7B). Despite the BMSC-promoted multiple myeloma cell growth, HDP-101 treatment showed even stronger cytotoxic effects (IC₅₀, 0.26 nmol/L) than in a no-BMSC microenvironment (IC₅₀, 1.2 nmol/L). Taken together, neither cell adhesion nor soluble factors like IL6 or IGF-1 impaired HDP-101 cytotoxicity.

The most successful treatments against pathogenetically heterogeneous diseases like multiple myeloma are drug combinations with different modes of action. We assessed HDP-101efficacy in combination with standard-of-care compounds (SOC), such as carfilzomib, dexamethasone, pomalidomide, doxorubicin, and melphalan, in the sensitive U266 and the partially resistant RPMI-8226 cell lines. Drug concentrations were selected according to their respective IC_12.5, IC₂₅, and IC₅₀ values of each drug and cell line, and combination indices (CI) were determined using the Chou–Talalay algorithm (refs. 33, 34; Supplementary Table S1).

Synergistic effects were observed between HDP-101 and carfilzomib, dexamethasone, pomalidomide, and melphalan in RPMI-8226 suggesting that combinations might overcome the partial resistance phenotype of this cell line (Fig. 7C). In contrast, doxorubicin or pomalidomide was antagonistic in U266 (Supplementary Table S1). These results suggest HDP-101 can enhance multiple myeloma cell sensitivity to several multiple myeloma SOC agents, and combination regimens with dexamethasone and carfilzomib/dexamethasone might be considered in clinical trial design.

Treatment of relapsed/refractory multiple myeloma remains a therapeutic challenge generally attributed to increasing clonal heterogeneity at relapse. The development of therapies using BCMA as a target for drug delivery represents a promising therapeutic approach for patients with multiple myeloma.

BCMA is expressed on malignant PCs but is absent on naïve B cells, germinal center B cells, or memory B cells making BCMA a selective target for multiple myeloma treatment compared with monoclonal antibodies against CD38, an antigen which is also expressed on normal cells (35). Moreover, BCMA expression is maintained through relapse after therapy (20, 36, 37). An anti-BCMA ADC conjugated to MMAF (GSK2857916) was shown to induce its cytotoxic effect dependent on the BCMA expression levels on multiple myeloma cell lines (38). We show HDP-101 binds BCMA with an affinity in the nanomolar range, and it is internalized by tumor cells and presents very low antigen-independent cytotoxicity. Furthermore, amanitin delivered by HDP-101 in pico- to nanomolar concentrations effectively induces cell death via apoptosis in multiple myeloma cells, sparing BCMA-negative cells. This occurs despite the BCMA low-level expression on some multiple myeloma cell lines and patient multiple myeloma cells. It, therefore, appears only small amounts of BCMA cell surface expression (<300 B.M. molecules/cells) are sufficient for HDP-101 to reach the required intracellular concentration of released amanitin to inhibit the RNA polymerase II complex resulting in multiple myeloma cell death.

Reported sBCMA levels in sera of patients with multiple myeloma correlate with clinical status and OS (19, 39). Recent studies suggest elevated sBCMA levels interfere with anti–BCMA antibody-based immunotherapy in multiple myeloma. Experiments using carefully titrated spiked sBCMA concentrations in this study indicated only very high sBCMA levels are likely to affect HDP-101 cytotoxicity. However, the highest sBCMA concentrations used in this assay were >2.5-fold more than sBCMA levels detected in multiple myeloma cell lines and >1.4-fold more than sBCMA levels reported in patients with multiple myeloma (40). This indicates that HDP-101 effectively induces multiple myeloma cell death even in the presence of sBCMA.

Importantly, we demonstrated that HDP-101 remains stable in the plasma for prolonged periods. The active compound bioavailability results in reliable in vivo performance. Moreover, an NCI-H929 s.c. xenograft model exhibited dose-dependent tumor regression and showed complete remission after a single i.v. treatment with HDP-101 at a dose of 4.4 mg/kg. In the disseminated MM.1S Luc xenograft model, a single dose of HDP-101 of 1 and 2 mg/kg i.v. resulted in durable tumor eradication with most mice remaining tumor-free for the observation period.

The Amanita phalloides mushroom toxin, amanitin, causes high mortality rates after ingestion due to liver failure and following acute renal toxicity (26). Mandatory HDP-101 full toxicity evaluation is necessary before introducing HDP-101 to humans. We performed toxicology studies in cynomolgus monkeys because of similar BCMA expression and cross-reactivity profiles to humans (41). HDP-101 was sequentially applied in doses of 0.3 or 1.0 mg/kg, and multiple-dose applications of 4 × 3.0 mg/kg. Amanitin is exclusively taken up by hepatocytes, and it is not expected to show toxicity on other cell types or tissues due to low membrane permeability. Transient mild-to-moderate increases in liver enzymes and lactate dehydrogenase were observed. These increases in liver enzymes are not considered toxic since histopathologic evaluation at the study did not show any signs of liver toxicity. With regard to nephrotoxicity, HDP-101produced dose-dependent pathology and/or clinical pathology findings in the kidney. However, findings of hyaline tubular casts, interstitial cortical fibrosis, and hyaline eosinophilic droplets in the kidney showed complete or partial recovery. This indicates reversibility of these findings. Safety profiling in cynomolgus monkeys revealed good tolerability and therapeutic index, likely due to very low levels of free amanitin.

We further show clear evidence of HDP-101 efficacy in primary multiple myeloma cells derived from both patients with newly diagnosed and relapsed/refractory multiple myeloma, irrespectively of the disease stage. It is known that multiple myeloma cells grow slowly in most patients and they do not proliferate in vitro, factors that contribute to their multidrug resistance phenotype. To our knowledge, there are currently three other ADCs against BCMA undergoing clinical evaluation in multiple myeloma, namely MEDI2228 (40), AMG 224 (42), and GSK2857916 (38). The MEDI2228 payload is pyrrolobenzodiazepine, a DNA cross-linker agent, and the AMG 224 payload is maytansine, a microtubule inhibitor; lastly, the GSK2857916 targets BCMA using MMAF as payload. Although those ADCs specifically target and effectively kill malignant PCs, like most other ADCs, the underlying mode of action of their respective payload is interference with microtubule function or cellular metabolism. In contrast, we here demonstrate that HDP-101 with amanitin's distinctive mode of action, as an RNA polymerase II inhibitor, does not interfere with the cell cycle and is independent of the ongoing cell proliferative status. Thus, it has potent cytotoxic effects on this notoriously hard-to-treat nonproliferating and dormant multiple myeloma cells, which are regarded as the main reason for disease spread to soft tissues and resistance (43, 44). This makes HDP-101 also a promising candidate for current clinical challenges such as minimal residual disease eradication.

We also investigated HDP-101efficacy in combination with other multiple myeloma therapeutic agents to find promising combination regimens for further clinical testing. We found a strong synergistic effect between HDP-101 and dexamethasone, likewise synergy with carfilzomib, in particular, thereby informing future clinical trial design.

To the best of our knowledge, no resistance mechanisms are so far known for amanitin in humans, and several attempts to induce in vitro resistance have failed. Besides, due to its hydrophilic nature, it is a null substrate to MDR-1 (45).

In conclusion, HDP-101, an ADC targeting BCMA with amanitin as the payload, shows potent cytotoxicity against both proliferating and resting BCMA-positive multiple myeloma cells in vitro and in vivo. Besides, HDP-101 was well-tolerated by cynomolgus monkeys in safety profiling. Thus, we have established HDP-101 merits further clinical assessment given its distinct mode of action and ability to overcome drug resistance.

V. Figueroa-Vazquez reports personal fees from Heidelberg Pharma Research GmbH during the conduct of the study. C. Breunig reports personal fees from Heidelberg Pharma AG during the conduct of the study; personal fees from Heidelberg Pharma Research GmbH; and personal fees from Heidelberg Pharma AG outside the submitted work. N. Giesen reports grants from Heidelberg Pharma GmbH during the conduct of the study; personal fees from MSD, Roche, and Pfizer; grants from BMS; and grants from Karyopharm outside the submitted work. A. Pálfi reports personal fees from Heidelberg Pharma Research GmbH during the conduct of the study; personal fees from Heidelberg Pharma Research GmbH; personal fees from Heidelberg Pharma AG outside the submitted work; and a patent for WO2018115466A1 issued to Heidelberg Pharma Research GmbH. C. Müller reports personal fees from Heidelberg Pharma Research GmbH during the conduct of the study; personal fees from Heidelberg Pharma Research GmbH; personal fees from Heidelberg Pharma AG outside the submitted work; and a patent for WO2018115466A1 issued to Heidelberg Pharma Research GmbH. C. Lutz reports personal fees from Heidelberg Pharma Research GmbH outside the submitted work and a patent for WO2018115466A1 issued to Heidelberg Pharma Research GmbH. T. Hechler reports personal fees from Heidelberg Pharma Research GmbH during the conduct of the study; personal fees from Heidelberg Pharma Research GmbH; personal fees from Heidelberg Pharma AG outside the submitted work; and a patent for WO2018115466 issued to Heidelberg Pharma Research GmbH. M. Kulke reports personal fees from Heidelberg Pharma Research GmbH during the conduct of the study; personal fees from Heidelberg Pharma Research GmbH; personal fees from Heidelberg Pharma AG outside the submitted work; and a patent for WO2018115466A1 issued to Heidelberg Pharma Research GmbH. H. Goldschmidt reports personal fees, nonfinancial support, and other from Amgen; grants, personal fees, and other from BMS; grants, personal fees, nonfinancial support, and other from Celgene; grants, personal fees, nonfinancial support, and other from Janssen; grants, personal fees, non financial support, and other from Chugai; grants, personal fees, nonfinancial support, and other from Sanofi; other from Takeda; personal fees and other from Novartis; other from Incyte; other from Molecular Partners; other from Merck Sharp and Dohme (MSD); other from Mundipharma during the conduct of the study; other from Adaptive Biotechnology; and personal fees from Glaxo Smith Kline (GSK) outside the submitted work. A. Pahl reports personal fees from Heidelberg Pharma Research GmbH during the conduct of the study; personal fees from Heidelberg Pharma Research GmbH outside the submitted work; and a patent for WO2018115466A1 pending and licensed to Heidelberg Pharma Research GmbH. M.S. Raab reports grants from Heidelberg Pharma Research GmbH during the conduct of the study; personal fees from GSK, non financial support from Takeda, grants and personal fees from Novartis, grants and personal fees from Amgen, personal fees from BMS, grants and personal fees from Sanofi, and personal fees from Janssen outside the submitted work. No disclosures were reported by the other authors.

V. Figueroa-Vazquez: Conceptualization, data curation, investigation, writing–original draft. J. Ko: Conceptualization, data curation, investigation, methodology, writing–original draft. C. Breunig: Data curation, formal analysis, validation, investigation, writing–review and editing. A. Baumann: Investigation, writing–review and editing. N. Giesen: Resources, writing–review and editing. A. Pálfi: Formal analysis, investigation, writing–review and editing. C. Müller: Data curation, formal analysis, investigation, writing–review and editing. C. Lutz: Formal analysis, investigation, writing–review and editing. T. Hechler: Conceptualization, formal analysis, investigation, writing–review and editing. M. Kulke: Data curation, methodology, writing–review and editing. C. Müller-Tidow: Supervision, writing–review and editing. A. Krämer: Resources, supervision, writing–review and editing. H. Goldschmidt: Resources, writing–review and editing. A. Pahl: Conceptualization, resources, supervision, funding acquisition, writing–original draft. M.S. Raab: Conceptualization, data curation, formal analysis, supervision, validation, methodology, writing–original draft, project administration.

The authors thank all patients who participated in this study. Part of this work was supported by grants from Heidelberg Pharma Research GmbH (D.10062038) and the Dietmar Hopp Foundation (1DH1911364).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.