Abstract
Background: In the context of precision diagnosis for various subtypes
of melanoma, identifying biomarkers with clinical translational
potential from a molecular standpoint is crucial for a more
comprehensive characterization of the disease. MUC18 is highly expressed
in both tumor cells and tumor vasculature in major melanoma subtypes and
is restricted to normal tissues.
Methods: A noninvasive imaging approach for MUC18 in melanoma utilizing
an immune Positron Emission Tomography (PET) radionuclide-conjugated
drug (RDC) with an 89Zr-labeled humanized anti-MUC18
monoclonal antibody (mAb) was developed. A375, Sk-Mel-28, HMVII, and
A549 cells and tumor model mice were conducted. Immuno-PET was employed
to assess the specificity and targeting of three distinct melanoma cell
line-derived xenografts (CDXs) and patient-derived tumor xenografts
(PDXs) in immunodeficient mice.
Results: The developed RDC, named 89Zr-IP150,
demonstrated robust in vitro stability and high binding affinity,
ensuring reliable and specific PET imaging of small, medium, and large
subcutaneous tumors in human melanoma mouse xenotransplantation models.
Notably, for the first time, the clinical translational potential of89Zr-IP150 was successfully validated using a PDX
model.
Conclusions: These findings present a noninvasive, real-time method for
the early screening of MUC18 (+) melanoma patients and are important for
studying the early-stage biological distribution of MUC18-targeted
antibody‒drug conjugates (ADCs).
Keywords: Immuno-PET imaging; Humanized mAb; Melanoma; PDX
model
Introduction:
Melanoma stands out as the most lethal form of skin cancer and
originates from the malignant transformation of
melanocytes[1]. These melanocytes, which arise
from the neuroectoderm, migrate extensively throughout the body,
including through the skin, uvea, mucous membranes, inner ear, and
rectum. These cells are highly dendritic cells that produce melanin to
shield against light damage[2]. According to
statistics, an estimated 100,640 individuals will receive a new
diagnosis of this invasive disease, and 8,290 individuals will die from
melanoma of the skin in America by 2024[3].
Benefiting from advancements in early clinical detection and systemic
treatment, the 5-year relative survival rate for skin melanoma patients
increased to 93% between 2011 and 2017, a notable improvement from 82%
in the mid-1970s[4]. Approximately 71% of
melanoma patients receive a diagnosis at stage 1, with a remarkable
5-year relative survival rate nearing 100%. In addition, although the
incidence of mucosal melanoma is low, due to the lack of early
identification, its mortality rate is much greater than that of skin
melanoma patients. Consequently, the early screening and detection of
melanoma holds profound significance in guiding subsequent treatment
choices and enhancing overall survival rates.
MUC18, also known as melanoma cell adhesion molecule (MCAM), is a
transmembrane glycoprotein identified by Johnson et al. using the
anti-human melanoma monoclonal antibody MUC18, which exhibits distinct
expression differences between malignant melanoma cells and normal
cells[5]. The resulting amino acid sequences
revealed that MUC18 belongs to the immunoglobulin superfamily and shares
significant similarity with the sequences of a group of nerve cell
adhesion molecules expressed during organogenesis[6,
7]. As research progressed, MUC18 acquired various names, including
CD146, melanoma adhesion molecule, melanoma-associated antigen A32,
melanoma-associated antigen Mel-CAM, MET-CAM, and HEMCAM. Through
bidirectional interactions with multiple specific ligands, such as
laminin 411 and 421, galectin-1 and -3, S100A8/A9, and matriptase, MUC18
actively participates in numerous physiological and pathological
cellular processes. Overexpression of MUC18 is commonly observed in most
malignancies and is implicated in nearly every stage of cancer
development and progression, particularly in vascular and lymphatic
metastasis[8]. In recent decades, immunotherapy,
including immune checkpoint blockade, has revolutionized cancer
treatment[9, 10]. However, T-cell exhaustion is
associated with decreased efficacy of immune checkpoint inhibitors and
adoptive T-cell therapies[11], and a significant
number of patients still do not benefit from current forms of
immunotherapy. Given the unusually high expression of MUC18 in a variety
of tumors and its various roles in reshaping the tumor microenvironment,
specific diagnosis and targeted therapies targeting MUC18 may overcome
this barrier[12].
Routine clinical methods, such as immunohistochemistry (IHC) and
fluorescence in situ hybridization (FISH), for analyzing biomarker
expression levels in tumors are invasive and lack real-time access to
biomarker expression throughout the body. Recent advancements in nuclear
medicine devices, such as whole-body positron emission
tomography/computed tomography (PET/CT), now enable the acquisition of
higher quality images while reducing patient and staff doses and
acquisition times [13-15].
Imaging targeting MUC18 currently relies predominantly on PET probes,
with radiolabeled precursors encompassing MUC18-targeted monoclonal
antibodies (mAbs), F(ab’)2 fragments, and
scFv[16]. To align with the precursors’ half-life,
nuclides such as 52Mn, 64Cu,68Ga, 89Zr, and99mTc are utilized for labeling. These probes have
been investigated in various cancer models, including models of
malignant melanoma[17, 18], brain
tumors[19, 20], lung cancer[21,
22], hepatocellular carcinoma[23], and breast
cancer[24, 25]. Despite the significance of these
studies, the antibodies YY146 and Fab’2 TsCD146 employed in the present
study are nonhumanized antibodies. In clinical applications, the human
body recognizes these monoclonal antibodies as alloproteins, leading to
immune rejection and faster clearance rates. Consequently, there is a
compelling need to develop humanized mAb-based radiopharmaceuticals
targeting MUC18 to address these challenges. We previously employed a
Zr-89-labeled humanized antibody for CLDN18.2-targeted
imaging[26].
Here, we developed an MUC18-specific probe through the radiolabeling of89Zr with a humanized monoclonal antibody, IP150. The
resulting radiolabeled probe, denoted RDC
(89Zr-IP150), demonstrated robust binding specificity
for MUC18 in vitro and exhibited reliable imaging capabilities for both
skin and mucosal CDX models in vivo. To enhance the clinical relevance
of this study, probe specificity was also assessed in melanoma PDX
models. These findings offer a promising approach for the early
diagnosis of melanoma patients, signifying the importance of exploring
the biological distribution of ADCs and guiding therapy.
Materials and Methods
General: 89Zr is produced using a medical cyclotron
via the nuclear reaction of 89Y (p, n)89Zr (Peking University Cancer Hospital & Institute).
MUC18 mAb (IP150, a humanized antibody) was generated in a CHO
expression system and purified by MabSelectTM Sure
Resin (GE Healthcare Life Sciences).