Jason S. Lewis
The Lewis lab research interests are focused on the development of new PET radiopharmaceuticals for the diagnosis and treatment of cancer. Our work incorporates 89Zr, 18F, 11C, and other nonstandard nuclides, with an emphasis on designing novel PET agents for cancer imaging. Our research program is a molecular-imaging-based program focused on the targeting of nonstandard nuclides as well as the development of small- and biomolecule-based agents and their clinical translation. A few examples of our work are given below.
89Zr-based Imaging Agents
One of the primary research focuses of our laboratory is the synthesis, development, and validation of 89Zr-based biomolecular imaging agents. One of the most fundamental principles in the construction of effective antibody-based nuclear imaging agents is matching the physical half-life of the radioisotope to the pharmacokinetic half-life of the targeting vector. This exigency has led our lab and others to seek to exploit the positron-emitting radiometal 89Zr (t1/2 = 78.41 h) as an effective and versatile long-lived radiolabel for PET imaging. Currently, our laboratory is pursuing a wide variety of projects that employ 89Zr as a radiolabel, though three imaging agents in particular combine to paint a representative picture of the breadth of our work: 89Zr-J591, 89Zr-transferrin, and 89Zr-5B1.
The first, 89Zr-J591, is a radiolabeled antibody that targets prostate-specific membrane antigen (PSMA), a well-characterized and important cell surface marker for prostate cancer. Our data reveal that 89Zr-J591 is highly effective at delineating PSMA-positive prostate cancer xenografts and, further, is capable of distinguishing between PSMA-positive and PSMA-negative tumors (Figure 1). The preclinical data was, in fact, so successful that 89Zr-J591 has been translated to the clinic where it is being employed in a phase I clinical trial.
The second agent, 89Zr-transferrin, offers a noninvasive technology for quantitatively measuring an oncogenic signaling pathway, more specifically the MYC status of tumors. The imaging agent binds the transferrin receptor 1 (TFRC, CD71) with high avidity, and thus can be used to produce high-contrast PET images that quantitatively reflect treatment-induced changes in MYC-regulated TFRC expression in a MYC-driven prostate cancer xenograft model (Figure 2). Indeed, the agent has been shown to detect the in situ development of prostate cancer in a transgenic MYC prostate cancer model, as well as in prostatic intraepithelial neoplasia (PIN) before histological or anatomic evidence of invasive cancer.
Finally, the most recent example is our work on the third agent, 89Zr-5B1. Serum circulating antigens are gaining attention as molecular determinants of cancer. One of these antigens is the CA19.9 or sialyl Lewisa, the primary biomarker of pancreatic ductal adenocarcinoma (PDAC). Currently, serological CA19.9 antigen assay is the standard clinical screening tool for detecting primary and metastatic disease and for evaluating response to therapy in PDAC. Other malignancies reported to have elevated levels of this antigen in the blood include those stemming from the colon, ovaries, and lung; this makes CA19.9 a target with broad utility across different cancers. Given these relationships, we have utilized a fully human, CA19.9-targeting monoclonal antibody, 5B1, as our vector and radiolabeled it with 89Zr in collaboration with MabVax Therapeutics, Inc., the inventor of this antibody. Through imaging and biodistribution studies, we have demonstrated the exquisite specificity of 89Zr-5B1 for its target (Figure 3).
Diels-Alder Click Chemistry
The remarkable specificity and affinity of antibodies make them extremely attractive vectors for the delivery of diagnostic and therapeutic radioisotopes to cancer cells. Over the past two decades, a wide variety of antibody-based radiopharmaceuticals have been successfully developed, employing isotopes ranging from 89Zr for imaging to 225Ac for therapy. A significant limitation of the use of radiolabeled antibodies, however, is that their relatively slow pharmacokinetics mandate the use of therapeutic isotopes with multiday physical half-lives. This combination of long biological and physical half-lives gives rise to an important complication: high activity concentrations in and radiation doses to nontarget organs.
In order to circumvent this problem, considerable attention has been dedicated to the development of targeting methodologies that combine the advantages of antibodies with the pharmacokinetics of smaller molecules. One particularly appealing method of achieving this balance while still employing intact antibodies is termed pretargeting. Generally, pretargeted methodologies involve four steps: (1) the injection into the bloodstream of a bivalent antibody with the ability to bind both an antigen and a radioligand; (2) the slow accumulation of the antibody in the tumor and concomitant clearance of the antibody from the blood; (3) the injection into the bloodstream of the small-molecule radioligand; and (4) the binding of the radioligand to the antibody followed by the rapid clearance of excess radioactivity.
One of the primary research foci of our laboratory is the development of pretargeting strategies based on an emergent type of click chemistry: the inverse electron demand [4+2] Diels-Alder cycloaddition between a tetrazine and a strained alkene dienophile (Figure 4). This reaction is rapid, selective, and, most importantly, bioorothogonal, and thus presents a near ideal chemical system for in vivo chemistry. To date, we have used this chemical technology to develop a modular strategy for the radiolabeling of antibodies(3)) and, more recently, to create an effective and robust pretargeted PET imaging strategy for colorectal cancer.(4)This system — which employs the colorectal cancer-targeting antibody A33 and the positron-emitting radiometal 64Cu — was shown to successfully image SW1222 colorectal cancer xenografts with tumor-to-background contrast comparable to antibodies directly labeled with 64Cu or 89Zr. Further, the rapid pharmacokinetics and extremely low nontarget tissue uptake of the tetrazine radioligands translates to background radiation doses that are only a fraction of those resulting from huA33 directly labeled with 89Zr.
Currently, our laboratory is actively working on expanding the radiochemical applications of Diels-Alder click chemistry, with projects dedicated to the development of pretargeted radioimmunotherapy methodologies, the creation of pretargeted imaging strategies using peptide-based vectors, and the use of this biorthogonal click chemistry with nanoparticle-based theranostic agents.
Dual-Modality PET/MR Longitudinal Imaging of Cancer Using Somatostatin Receptor–Targeted Liposomes
The emergence of dual modality PET/MR scanners in the clinic has made possible a new dimension of diagnostic imaging: high-resolution, functional, anatomical imaging with minimal ionizing radiation dose to the patient. Yet despite the advent of PET/MR technology, very few dedicated dual-modality imaging agents have been developed to fully exploit its potential. The development of such imaging agents has become a principal research focus in our laboratory.
In this regard, we have recently reported a dual-modality pegylated liposome formulation for simultaneous PET and MR imaging bearing an MR T1 contrast agent and the positron-emitting radiometal 89Zr. (Figure 5, top). These pegylated liposomes feature a chelator-free 89Zr-radiolabeling strategy and, to achieve selectivity for tumors expressing the human somatostatin receptor subtype 2 (SSTr2), are conjugated to the octreotide peptide.
At 50 hours postinjection in a murine model of neuroendorcine cancer, the 89Zr-labeled, octreotide-conjugated liposomes were shown to selectivity target SSTr2-positive-overexpressing tumors as shown by coregistration PET and MR images. In sum, we have presented that multifunctional cargo liposomes show long systemic circulation, SSTr2-targeting specificity, and dedicated dual PET/MR imaging capability (Figure 5, bottom).
Radiolabeled pH (Low) Insertion Peptide (pHLIP)
The microenvironment of solid tumors is often characterized by a phenomenon known as the Warburg effect, an increased reliance on glycolytic respiratory pathways that results in the over-production of lactic acid. This excess lactic acid in turn requires extracellular transport to sustain tumor survival; however, the poorly formed vasculature of tumors hampers the subsequent diffusion of this acidity. As a result, the extracellular milieu in malignant tissues is relatively acidic (pHe~6.5–6.8) compared to that of normal cells (pHe~7.3-7.4). This reduced pH, often termed tumor acidosis, is an extremely promising target for noninvasive cancer imaging.
In partnership with Yana Reshetnyak and Oleg Andreev at the University of Rhode Island and Donald Engelman at Yale University, our laboratory is focused on developing a PET imaging agent for tumor acidosis based on a peptide carrier known as “pH (Low) Insertion Peptide” (pHLIP). pHLIP is composed of a 22-37 amino acid sequence derived from the C-helix of Bacteriorhodopsin. As shown in Figure 6, in the acidic extracellular environment of a tumor, the peptide inserts in the cellular membrane, forming an α-helix and acting in essence as a nanosyringe. We have attempted to harness this peptide technology for imaging by conjugating radiometal chelates (68Ga, 64Cu) and 18F-fluorinated derivatives to various pHLIP variants and investigating the pharmacokinetic profile of these imaging agents in murine models of prostate cancer. The first generation pHLIP peptide is characterized by prolonged blood circulation, so in addition to working with this original variant, we have also continued to modify and improve the pharmacokinetics of pHLIP with various truncated sequences, radionuclide, and chelates.
To confirm proof of principle, we have partnered with Jason Koutcher and his group here at Memorial Sloan-Kettering to measure global and regional extracellular pH (pHe) in tumor xenografts via 31P/1H magnetic resonance spectroscopy in order to correlate tumor pHe and levels of pHLIP-PET uptake. We are also further investigating the targeting mechanism of this peptide through ex vivo digital autoradiography and histology studies.
- Holland, et al. 89Zr-DFO-J591 for ImmunoPET of Prostate-Specific Membrane Antigen Expression In Vivo. J. Nucl. Med. 2010; 51(8): 1293-1300. © The Society of Nuclear Medicine, Inc.
- Holland, et al. Annotating MYC Status with 89Zr-Transferrin Imaging. Nature Medicine 2012; 18(10): 1586-1592. © Nature Publishing Group, Inc.
- Zeglis, et al. Modular strategy for the construction of radiometalated antibodies for positron emission tomography based on inverse electron demand Diels-Alder click chemistry. Bioconjugate Chem., 2011 Oct 19;22(10):2048-59.
- Zeglis, et al. J. Nucl. Med., 2013