by Jennifer van Brunt

The discovery by Kohler and Milstein in 1975 that it was possible to generate monoclonal antibodies via hybridoma technology sparked a scientific revolution – and the consequences have been awesome. Not only has the ability to create endless supplies of specifically tailored, pure antibody molecules changed the face of scientific inquiry into the considerable complexities of the immune system per se, but also it has opened up an entirely new approach to disease therapy. Indeed: The FDA has approved 17 monoclonal antibody-based drugs to date – and hundreds more are in the clinic. Among the approved products are a number of highly successful cancer therapies, but clinical experience has proven that these monoclonals work best when dosed concurrently with or following chemotherapy. So-called "naked" antibodies are just not that good at killing cancerous cells on their own. That's why there has been a resurgence of interest in the use of "armed" antibodies – molecules that are linked to or fused with chemotherapeutic drugs, lethal toxin molecules or powerful radionuclides – for new cancer treatments. And companies developing arming technologies suddenly find themselves in great demand.

You had to have been an immunologist in 1975 to understand just how earthshaking it was. When Georges Kohler and Cesar Milstein published their landmark paper in Nature describing how to generate monoclonal antibodies via hybridoma technology, it completely changed the field. For researchers intent on unraveling the considerable complexity of the immune system suddenly had a dazzling new tool – and the possibilities seemed infinite.

Now they could create endless, and consistent, supplies of monoclonals to just about antigen on earth, rare or common. Having vats full of homogeneous molecules, it then became possible to run all sorts of assays and tests – without exhausting the original supply. That meant researchers could dissect the immune system at will and – with the help of recombinant DNA technology, which was coming into its own at the time -- they could start characterizing and isolating immune system genes. They could also analyze the structures of antibody molecules – and, eventually, determine ways to tinker with them to improve their antigen-binding capabilities.

Kohler and Milstein's Nobel Prize-winning technology created a revolution in the lab – but it also turned into a revolution in healthcare, vastly improving the precision of diagnostic tests (where would clinical labs be without enzyme-linked immunoassays?) and opening up an entire new approach to disease therapy. Indeed, the FDA has now approved 17 monoclonal antibody-based therapies – two of them just last month -- and hundreds more are in development to treat a wide range of ills, including immune disorders, infectious diseases and cancer. (The eight approved products for treating cancer are detailed in the tables below.)

But getting to this point hasn't been easy. Although it didn't take long for the first therapeutic to hit the market – the FDA approved Ortho Biotech Inc.'s Orthoclone OKT3 for reversing acute kidney transplant rejection in 1986 – the second product – Centocor Inc.'s ReoPro for high-risk coronary angioplasty – wasn't approved until 1994. The years in between were littered with clinical trial failures – so many that at one point the goal of harnessing the power of antibodies as therapeutic agents seemed hopeless.

FDA-approved monoclonal antibody therapies for cancer

In cancer, especially, it turns out that monoclonals are such large molecules that often they aren't able to penetrate tumors. Moreover, not all monoclonals have the ability to actually kill cells on their own. The first-generation products were composed of mouse proteins, which unfortunately caused immune reactions when injected into humans. That was deemed acceptable as long as the therapy was only given once – as with Orthoclone OKT3, which is a murine antibody – but this immune reaction totally precluded repeat administration. As well, these neutralizing antibodies could soak up enough of the medicine as to minimize its therapeutic benefits. On top of these challenges, it was very expensive to even make enough product for therapeutic use.

But all that's changed now, as researchers have come up with ways to circumvent the mouse problem – either by replacing about two-thirds (chimeric) or most (humanized) of the murine sequences with human ones or by producing fully human antibodies in transgenic mice or via phage display technologies. Scientists have figured out the antibody targets, the correct antibodies to be used in various treatment scenarios and the correct treatment regimens. Production methods have improved, too, allowing companies to make large quantities of monoclonals in a cost-effective manner.

Still, it would be premature to state that the course of antibody product development now runs smooth. For many hurdles remain – particularly in cancer, where the challenge is to increase an antibody's ability to kill the cells it's targeting. And new technologies under development, as well as improvements on older approaches, are paving the way.

Some antibodies, of course, are able to directly kill cells by inducing apoptosis. But not all exert their effects in such a straightforward manner. Some block various growth factors, thereby halting the proliferation of tumor cells. Others work indirectly, by recruiting cytotoxic cells such as macrophages and monocytes (a phenomenon called antibody-dependent cell-mediated cytotoxicity; ADCC) or by binding to complement, which is also a cytotoxic event (a phenomenon known as complement dependent cytotoxicity; CDC). Some probably use more than one mechanism.

For instance, Rituxan, developed by IDEC Pharmaceuticals Corp. and Genentech Inc. and the very first monoclonal antibody approved in the U.S. for treating cancer, binds specifically to the CD20 antigen expressed on the surface of B cells and induces their lysis. It's been proposed that this occurs through ADCC and/or CDC, but the exact mechanism is still not known.

And Herceptin, which targets human epidermal growth factor receptor 2, might block the signaling pathway that triggers further cell growth by causing some of those receptors to be endocytosed (a cytostatic mechanism). It's also been proposed that Herceptin treatment prevents DNA repair following a round of DNA-damaging chemotherapy (cytotoxic). And, the antibody could also work via ADCC, attracting natural killer cells that attack and consume the cancer cells. Here again, the exact mechanism of action is not fully understood.

FDA-approved monoclonal antibody therapies for cancer

But what is known is this: The median survival of patients with HER-2 overexpressing metastatic breast cancer is better when they are treated with Herceptin plus paclitaxel than it is when they receive chemotherapy alone.

And, in fact, the monoclonals approved for use in treating cancer are routinely prescribed either concurrently with or following chemotherapy. It's a far cry from the early days, when researchers thought that antibodies would be able to do the job all on their own.

For instance, Genentech's Herceptin is prescribed as a first-line therapy only in combination with chemotherapy. It can, however, be used as a second- or third-line therapy as a single agent. And the company's newly-approved drug Avastin is prescribed as a first-line therapy, in combination with 5-fluorouracil, for patients with colorectal cancer.

ImClone Systems Inc.'s colorectal cancer therapy Erbitux is prescribed as a combination therapy with irinotecan in patients who are refractory to irinotecan or as a single agent in patients who can't tolerate the chemotherapy. Ilex Oncology Inc.'s Campath for B-cell chronic lymphocytic leukemia is dosed only after patients have been treated with alkylating agents and have failed fludabarine therapy.

Since combining antibody therapy with chemotherapy works in cancer, why not fuse a chemotherapeutic agent directly to an antibody molecule? That way, the antibody could deliver the toxic agent directly to the tumor site and spare normal cells from its devastating effects. Well, Wyeth's monoclonal drug Mylotarg, an anti-CD33 monoclonal conjugated with calicheamicin, was designed to do precisely that. Approved in 2000 for treating relapsed acute myeloid leukemia in elderly patients, Mylotarg is the first – and only – antibody-drug conjugate on the market.

It's also possible to attach a radionuclide to an antibody molecule – arming it to deliver intense radiation directly to the cancerous growth. Biogen Idec Inc.'s Zevalin was the first radioimmunotherapy to garner FDA approval; a year or so later, Corixa Corp.'s Bexxar followed suit. Both products are prescribed for treating non-Hodgkin's lymphoma in patients whose disease has relapsed and is refractory to Rituxan. Both are used as part of a multi-step treatment regimen. Bexxar is administered in two steps – dosimetric (which assesses biodistribution of the antibody to make sure it's on target) and therapeutic -- each of which consists of a sequential infusion of unlabeled antibody followed by I-131 labeled antibody. Zevalin's therapeutic regimen is similar in design, but involves the use of two different isotopes – Indium-111 for imaging (biodistribution) and Yttrium-90 for the therapeutic dose.

There's no doubt that armed antibodies work – and the three already on the market represent important alternatives to the "naked" antibody-based cancer therapies now available. But using antibodies to deliver payloads of toxins, radionuclides or chemotherapeutic drugs is not a new concept. In fact, researchers have been trying to achieve this goal for decades. Why has it taken so long? Frankly, they've had as much trouble understanding how to construct effective drugs of this type as they did with naked antibodies. Issues surrounding toxicity, linker technology and manufacturing, among others, had to be worked out first. But the tide is turning – and interest in exploiting the potential of armed antibodies is on the rise.

That's the main reason that Seattle Genetics Inc. has signed so many partnerships of late: The firm, founded in 1997 and a relative newcomer to the playing field, not only develops "naked" antibodies as therapeutics but also had devised one technology that creates antibody-drug conjugates (ADC) and another that is used to make antibody-directed enzyme prodrug therapies (ADEPT).

"Drug conjugates can empower an antibody," explained Clay Siegall, the company's president and CEO. The vast majority of antibodies are just not very potent on their own, he added. For example, "Herceptin is best when used in combination [with a drug]." There are other antibodies now available for treating cancer "that can be used on their own," he said, "but they are most effective when used with a series of toxins."

Seattle Genetics' second-generation ADC technology, which is still in early development, uses stable linkers to join antibody molecules with highly potent cell-killing drugs. The company uses this technology in its own preclinical product candidate SGN-35 and has out-licensed it to partners Eos Biotechnology Inc. (since acquired by Protein Design Labs Inc.), Genentech and Celltech Group plc for use with each firm's proprietary monoclonal antibodies.

Armed antibodies under development for treating cancer

To make these ADC constructs, Seattle Genetics chooses antibodies that internalize upon binding to their cell-surface receptors. In cancer, "the antibodies have to have the ability to traffic into the cell, to internalize," Siegall said. "The more rapid the internalization, the better the antibodies work." Once inside the cell, they are taken up by lysosomes, where the linker is cleaved and the drug is released.

The linkers that join the drug to the antibody are vitally important to its efficacy. Seattle Genetics uses linkers that are stable outside the cell, so they cannot be broken prematurely and allow the drug to float around in the circulation. The first-generation linkers are pH dependent; the second-generation linkers are enzyme-cleavable. "In the past, people have used linker systems that weren't stable," Siegall explained. "If they're not stable, the drug falls off in the blood. Our enzyme-cleavable linker is active inside tumor cells but not in the blood."

Moreover, the drugs that Seattle Genetics has chosen to use are variants of Auristatin E, a very potent cell-killer that blocks DNA replication. Importantly, these drugs are inactive until they are released inside the target cells, thus sparing normal cells. Because the compounds are synthetically produced, scale-up is straightforward. "In the past, people used drugs off the shelf and all of them were natural products," Siegall said. "We use a synthetic drug, so it is easy to make."

Seattle Genetics' lead ADC product candidate SGN-15, which is in Phase II trials in non-small cell lung cancer, is a first-generation product composed of the chimeric BR96 antibody chemically joined by an acid-labile hydrazone linker to the chemotherapeutic drug doxorubicin. The antibody component binds to a Lewisy—related carbohydrate antigen that is highly expressed on many solid tumors; it internalizes rapidly and then releases the drug payload at the low pH present with the tumor cells. SGN-35, which uses the next generation ADC technology, is in preclinical development. It consists of an anti-CD30 monoclonal attached to a variant of Auristatin E through a protease-sensitive, synthetic dipeptide linker. This antibody is also internalized rapidly, Siegall said; once inside the lysosomes, the linker is cleaved by capthepsin B and drug is released.

While the firm's ADC technology uses internalizing antibodies, its ADEPT technology specifically uses non-internalizing ones. In the case of SGN-17/19, which Seattle Genetics is developing with partner Genencor International Inc. for treating metastatic melanoma, the product concentrates the potent cell-killing properties of the drug melphalan towards cells expressing the p97 antigen. To achieve this, it uses the cloned variable regions of a non-internalizing monoclonal antibody genetically fused to an enzyme; once administered to the patient, this product accumulates on the tumor's surface. In a second step, the relatively inactive prodrug is administered and subsequently converted by the enzyme attached to the tumor cell into an active drug that penetrates the tumor tissue.

Thus, Seattle Genetics has come up with two clever ways to hone in on cancer cells specifically and deliver the deathblow from within.

But there are other variations on this theme, as well. Biotech and big pharma companies are also flocking to license the new technology developed by ImmunoGen Inc. – one of the very first biotechs to explore the potential of armed antibodies as therapeutic agents.

Back in the mid-1980s, ImmunoGen's first weapon of choice was ricin, extracted from the castor bean and the most toxic substance known to man. The toxin molecule consists of two chains, with the A chain being the killer: One ricin A chain can inactivate about 2,000 ribosomes per minute, thereby rapidly disrupting protein synthesis.

The company developed a number of product candidates that linked blocked ricin to various antibodies and took four of them into clinical trials. But when it restructured in 1994, it stopped development of all but its lead candidate Oncolysin B, an anti-B4 antibody which was already in Phase III trials in B-cell lymphoma patients following autologous bone marrow transplantation (as well as earlier-stage trials in other indications). That trial never panned out, however – the preliminary data indicated that Oncolysin B offered no advantage over the control – and by the spring of 1997 ImmunoGen had ended its efforts to continue down this path.

According to Walter Blättler, ImmunoGen's executive VP of science and technology, "We found out that toxins from plants or bacteria were toxic and that murine antibodies can elicit an immune response. After 8-9 days, even the lymphoma patients had an immune response." As well, "the [ricin] toxin caused capillary leak syndrome; this was an unexpected toxicity and limited the dose."

In the meantime, the Cambridge, MA company was working on another approach – which it now calls tumor activated prodrug (TAP) technology. This time, it attached the chemotherapeutic agent DM1 (a small molecule derivative of mertansine, a natural product and a very potent inhibitor of cell division) to an antibody that binds to a protein on the surface of tumor cells. The linker that joins the two is stable in the circulation but is readily broken once it is internalized. "This small molecule [DM1] is non-immunogenic. It's as active as ricin but less toxic," Blättler explained.

In the blocked ricin constructs, "the antibody/antigen complex was internalized into the cell. The A chain was released from the B chain and inactivated the ribosomes. TAP exploits a similar internalization mechanism," he said. "The drug is released inside the cell. The target for maytansinoids [like DM1] is tubulin."

Company scientists also experimented with the cleavage properties of the linkers – for the most efficient and rapid release of the drug inside a cell -- and settled on sterically hindered disulfide bonds, which are relatively resistant to reduction in aqueous solution (meaning the product would have a relatively long "shelf life"). Once inside the cell, though, "glutathione reduces the disulfide bond" and releases the drug," Blättler said.

ImmunoGen is currently developing cantuzumab mertansine (a humanized antibody that targets CanAg, which is found on a number of cancers) linked to DM1, which has been evaluated in three Phase I clinical trials in patients with colorectal, pancreatic and lung cancer and should enter Phase II studies next year.

Armed antibodies under development for treating cancer

It's also testing huN901-DM1 in small-cell lung cancer (SCLC). This product consists of a humanized antibody that targets the CD56 antigen found on various types of tumors, conjugated with DM1. ImmunoGen's partner British Biotech plc (which merged with Vernalis Group plc in September 2003) was leading clinical development efforts – and will complete its U.K.-based Phase I study and continue its U.S.-based Phase I/II study until this June – but recently returned product rights to ImmunoGen, which will take over the product's future development.

In July 2003, ImmunoGen signed a broad-reaching oncology collaboration with Aventis Pharmaceuticals Inc. that gives Aventis rights to three early-stage products in the biotech's pipeline and also teams the partners in discovery efforts for new products. The committed research funding this deal provides allowed the biotech firm to advance cantuzumab mertansine and huN901-DM1 on its own.

Other firms are interested in the technology, too: ImmunoGen has granted single-target licenses on its TAP technology to Genentech (for Herceptin), Millennium Pharmaceuticals Inc. (for MLN2704, now in Phase I/II trials for metastatic prostate cancer), Boehringer Ingelheim International GmbH (for bivatuzumab mertansine, in Phase I trials for squamous cell carcinoma), and Abgenix Inc.

"All cancer companies are interested in antibodies," Blättler said. Not only are there eight antibody-based products already on the market, but also it's now clear that they can achieve blockbuster status: Rituxan sales reached nearly $1.5 billion in 2003.

"We have the antibodies," he said, "and now we have to make them better. We have to help the antibodies kill cells."

Founded in 1982, the year after ImmunoGen got its start, Immunomedics Inc. also focused on developing a new type of antibody-based cancer therapy – but instead of using the molecules to deliver toxic drugs, the firm chose to load them up with radionuclides.

In the early 1990s, Immunomedics tested several drug candidates in the clinic – all of them murine monoclonals labeled with I-131. Its most advanced product, ImmuRAIT-LL2, made it through Phase I/II trials in B-cell lymphoma patients, trials that were designed to provide information on antibody targeting and dosing. But this was a mouse antibody, which couldn't be used for repeat administration, so the company decided to halt that product's development and switch to humanized antibodies going forward. It also changed radionuclides, choosing yttrium-90 because it provides a superior level of high-energy radiation for killing cancer cells. As well, the firm has developed a method of conjugating Y-90 to the antibodies via the macrocyclic chelate DOTA, which results in extremely stable products that are resistant to dissociation.

"The conjugation technology is dramatically improved," explained Ivan Horak, Immunomedics' executive VP of R&D and chief scientific officer. "Less of the yttrium leaks after an injection when it is conjugated with DOTA rather than DTPA [the chelator used in Zevalin]," he said.

Armed antibodies under development for treating cancer

However, Immunomedics is still pursuing I-131-labeled compounds, thanks to its development of a new, residualizing labeling method that allows a 10-fold increase in the isotope's uptake and retention by cancer cells. Results of a Phase II trial using an I-131-labeled, humanized antibody against carcinoembryonic antigen in colorectal cancer patients who had their liver metastases removed were reported in late January at a GI Cancer Symposium.

"I-131 is more appropriate in an adjuvant setting or small volume tumors, where there is minimal residual disease or occult disease," Horak said. "Y-90 is harsher, and should be used in patients with clearly detectable lesions. It has better penetration."

Currently, the Morris Plains, NJ firm has two clinical-stage armed antibody drug candidates labeled with Y-90: LymphoCide Y-90, a humanized antibody targeting the CD22 receptor on B-cell lymphomas which is in Phase I trials for treating indolent and aggressive non-Hodgkin's lymphoma; and CEA-Cide Y-90, a humanized antibody targeting CEA (carcinoembryonic antigen)-expressing tumors which is in Phase I trials for treating inoperable metastatic solid tumors.

The company's majority-owned subsidiary IBC Pharmaceuticals Inc. is working on a different sort of approach that involves bispecific antibody, pretargeting technology to deliver a punch to tumor cells. In this two-step method, a humanized bispecific antibody, with one arm that recognizes a tumor-associated antigen and another that recognizes an epitope on a therapeutic agent, is given as a first injection. When the product has cleared non-target tissues and reached a maximum level in the tumor, the therapeutic agent – a chemotherapy drug, ribonuclease, or enzyme prodrug, linked to a peptide -- is dosed. This technology attracted the interest of Schering AG, which initiated a research program with IBC Pharmaceuticals in October 2003 to evaluate and potentially develop a cancer drug involving a Schering antibody.

There's yet one more "first-generation" company that's developed a second-generation pretargeting technology for delivering a payload to cancer cells. NeoRx Corp. created a platform that delivers intense doses of radionuclides directly to tumor cells by separating the targeting agent from the radiotherapeutic agent. In this multi-step process, a recombinant fusion protein consisting of an antibody fragment specific to the tumor of interest fused to strepavidin is administered, followed by a synthetic clearing agent to remove the unbound protein. Finally, a small molecule radiotherapeutic linked to biotin that attaches directly to the pre-localized fusion protein through a biotin-strepavidin recognition event is added.

NeoRx learned a lot from its first-generation product Avicidin. For one, the antibody employed was not tissue-specific enough, explained Mark Jones, the firm's senior director of R&D. It targeted the pan-carcinoma antigen EpCAM, which is not only expressed on tumors, but also on the epithelial cells that line the gut. In the clinic, the product exhibited gastrointestinal toxicity. As well, the first-generation clearing agent was based on albumin – which these days is universally shunned, along with any products derived from blood. However, Jones remarked, "Avicidin taught us that the [pretargeting] concept would work."

Using its refined technology, the Seattle-based company completed Phase I trials for two product candidates – in non-Hodgkin's lymphoma and colorectal carcinoma. But NeoRx has changed course, and now focuses on skeletal targeted radiotherapy (which uses a small molecule). Thus, the firm does not plan to develop its antibody-based products further, although it is seeking to out-license the rights to the products and the platform technology – which can be adapted for use with different antibodies and fusion proteins.

Armed antibodies under development for treating cancer

Ira Pastan's group at the National Cancer Institute (NCI) devised yet another method to target tumors with lethal agents. The scientists modified Pseudomonas exotoxin A, an extremely potent bacterial toxin, by deleting one of its three domains (the binding domain) and replacing it with the Fv fragment of an antibody molecule. One of the NCI's recombinant toxins, SS1 (dsFv)-PE38 (a.k.a. SS1P), targets mesothelin, a cell surface antigen overexpressed in mesothelioma, ovarian and pancreatic cancers.

NeoPharm Inc. licensed the rights to SS1P from the NCI and also entered into a collaborative research and development agreement (CRADA) in 1999 for the product's development and commercialization. NeoPharm took the product through Phase I clinical trials in patients with refractory mesothelin-modulated cancers – only to discover that all patients except one developed antibodies to the drug after one course of treatment. By late 2002, the Lake Forest, IL company ended its participation in the program and returned the rights to NCI, stating that SS1P did not meet its requirements for advancement into Phase II trials.

About a year later, Enzon Pharmaceuticals Inc. picked up the product, signing a CRADA with the NCI to continue the development of SS1P in pancreatic and ovarian cancer. The NCI is still conducting Phase I trials, and Phase IIs are set to start in the second half of this year. As part of the CRADA, Enzon will also use its own PEGylation technology to potentially improve the immunotoxin.

Armed antibodies under development for treating cancer

According to Ulrich Grau, Enzon's chief scientific officer, the antibodies raised in the patients who participated in NeoPharm's trials "were presumably to the exotoxin portion" of the immunotoxin product, not the antibody fragment. Enzon intends to explore "different dosing regimens" to ascertain whether it's possible to resolve this problem.

In parallel, however, the company will PEGylate the immunotoxin to see if that process will reduce its immunogenicity. "There is evidence from other proteins that modifying them with PEG elicits a reduced immune response," Grau explained. "We also know that attaching PEG residues passively targets proteins to the tumor site. The vascular bed in a tumor is leaky, and these molecules accumulate there and get trapped." And PEGylation has another benefit, too: It prolongs the product's half-life, meaning that tumors would get greater exposure to the immunotoxin molecule.

Seattle Genetics, ImmunoGen, Immunomedics and Enzon Pharmaceuticals are certainly not the only biotechs exploring the potential of armed antibodies – they're joined by Antisoma plc, Peregrine Pharmaceuticals Inc., Millennium Pharmaceuticals, Genentech, Abgenix, Protein Design Labs, Celltech, Genencor and no doubt others. And pharmaceutical companies – including Schering AG, Aventis and Boehringer Ingelheim -- have jumped on the bandwagon, too.

Using monoclonal antibodies to deliver killer payloads to tumors was a fresh and exciting idea in the 1980s – but a series of early trials that met with failure seemed to spell doom for the concept. Today, however, it's made a comeback, as compelling clinical evidence has demonstrated that most monoclonals really don't do a very good job at killing cancerous cells on their own. To really pack a wallop, antibodies need some help – and arming them with anticancer drugs, lethal toxin molecules or powerful radionuclides is suddenly the hottest new approach to cancer therapy.

What goes around comes around.

Editor's note: By sheer coincidence, the title of this article, "Armed Antibodies," is also a trademark owned by Viventia Biotech Inc.

Copyright © 2004. Signals is an online magazine of analysis for biotechnology executives. To contact the Signals editorial department, send an email. Signals is published by: Recombinant Capital, 2033 N Main Street, Suite 1050 , Walnut Creek, California 94596-3722, Phone: +1 (925) 952-3870

To make any comments on this article, or to ask a question of the author, please contact the publisher. If you would like to submit an article please subscribe to our PL Intelligence service.

The opinions expressed in the articles published in this section do not necessarily reflect those of Pharmalicensing or UTEK Corporation. No actions including proposals to or agreements with other companies should be taken by any reader without obtaining specific business or legal advice. Neither the publisher nor the authors accept any liability for any actions or activities undertaken by any reader or other third party as a consequence of these articles or for any errors or omissions therein.