Business

Overview

We discover and develop monoclonal antibody-based drugs to treat cancer and related diseases. We have four monoclonal antibody-based technologies: monoclonal antibodies; monoclonal antibodies chemically linked to cell-killing drugs, or monoclonal antibody-drug conjugates; single proteins containing monoclonal antibody and toxin components, or single-chain immunotoxins; and antibody-directed enzyme prodrug therapy, or ADEPT. Our technologies enable us to increase the potency and efficacy of monoclonal antibodies that have specificity for cancer or related diseases but are not potent enough on their own. Using our expertise in cancer and monoclonal antibody technologies, we have constructed a diverse portfolio of drug candidates targeted to many types of human cancer. Our four technologies provide us with the means to discover and develop monoclonal antibody-based drug candidates internally as well as to partner our technology with other companies also developing monoclonal antibodies.

We are testing our two most advanced drug candidates, SGN-15 and SGN-10, in patients with breast, colon, prostate or other solid tumor cancers. SGN-15 is a monoclonal antibody-drug conjugate that binds to cancer cells and kills them by delivering the drug doxorubicin inside the cell. We are currently testing SGN-15 in three phase II clinical trials in combination with the chemotherapy drug Taxotere. SGN-10 is a single-chain immunotoxin that binds to cancer cells and kills them by delivering a protein toxin inside the cell. We are testing SGN-10 in two phase I clinical trials, one as a single agent and the other in combination with Taxotere. Aventis is co-funding two of the SGN-15 clinical trials and one SGN-10 clinical trial.

We also have five drug candidates in preclinical development for the treatment of patients with solid tumors, melanoma or blood-cell cancers, commonly known as hematologic malignancies, including multiple myeloma and lymphomas. SGN-14 is our humanized monoclonal antibody targeted to the receptor identified as CD40, which we have licensed to Genentech for the development of therapies to treat patients with blood-cell cancers, commonly known as hematologic malignancies, or other types of cancer. SGN-30 is our monoclonal antibody targeted to the receptor identified as CD30, which we are developing for treatment of patients with hematologic malignancies and other types of disease. We are developing SGN-17/19, which utilizes our ADEPT technology, for treatment of patients with melanoma. We are developing two additional drug candidates that utilize our novel monoclonal antibody-drug conjugate technology, specifically our stable linkers and our proprietary, high-potency cell-killing drugs.

Monoclonal Antibodies as Therapeutics

Numerous monoclonal antibodies have been approved for treatment of autoimmune disease, cancer and infectious disease, and represent an important area of novel therapeutic product development. Antibodies are protective proteins released by the immune system's B cells, a type of white blood cell, in response to the presence of a foreign substance in the body, such as a virus. B cells produce millions of different kinds of antibodies with slightly different shapes that enable them to bind to and thereby affect different targets. Antibodies of identical molecular structure that bind to the same target are called monoclonal antibodies. Typically, mice have been used to produce monoclonal antibodies to a wide variety of molecular targets, including targets to which the human body does not normally produce antibodies. In particular, many mouse monoclonal antibodies have been developed as potential therapeutics to neutralize viruses, destroy cancer cells or inhibit immune function.

Although mouse monoclonal antibodies are relatively easy to generate, they have significant drawbacks as therapeutics. Mouse monoclonal antibodies have a relatively short half-life in human patients, requiring them to be administered frequently. Moreover, mouse monoclonal antibodies are not adapted to work effectively with the human immune system and therefore often have limited ability to destroy the target, such as cancer cells. Most importantly, when injected into human patients, a mouse monoclonal antibody is usually recognized by the body's immune system as being foreign. The immune system therefore responds

26

by rapidly neutralizing the mouse monoclonal antibody and rendering it ineffective for further therapy. These problems have largely prevented mouse monoclonal antibodies from becoming therapeutics.

Recognizing the limitations of mouse monoclonal antibodies, researchers have developed a number of approaches to make them appear more human-like to a patient's immune system. For example, improved forms of mouse monoclonal antibodies, referred to as "chimeric" and "humanized" antibodies, are genetically engineered and assembled from portions of mouse and human antibody fragments. Chimeric antibodies contain approximately 35% mouse sequences and humanized antibodies contain approximately 15% mouse sequences. Additionally, monoclonal antibodies have also been prepared in fully human form. These technologies have enabled scientists to develop antibody products that can be administered to patients repeatedly over time, or on a chronic basis, with reduced adverse responses by the human immune system. Similarly, advances in monoclonal antibody production technologies, such as high productivity fermentation and transgenic plants and animals, have enabled manufacturers to produce monoclonal antibody-based products more cost-effectively. Because of these advances, a large number of monoclonal antibodies are currently undergoing clinical and preclinical investigation. According to a survey conducted by the Pharmaceutical Research and Manufacturers of America, 72 out of 369, or 20% of all biotechnology medicines in clinical trials in 1999 were antibodies. The FDA has approved nine therapeutic antibodies, seven of them in the last three years, with total sales in 1999 in excess of $1.4 billion worldwide.

Monoclonal Antibodies for Cancer Therapy

Cancer is the second leading cause of death in the United States, resulting in over 550,000 deaths annually. The National Cancer Institute reports that more than 8 million people in the United States have cancer and they estimate that one in three Americans will develop cancer in their lifetimes. Approximately 1.2 million new cases of cancer were diagnosed in 2000 in the United States. Although there are many commercially available products to treat various forms of cancer, most are not curative. Even small improvements in therapies represent precious time for patients dying from cancer, and for many patients, there are no meaningful therapies available.

Monoclonal antibodies have been tested for many years as cancer therapeutics. However, while some monoclonal antibodies have significant antitumor activity as single agents, many are not potent enough on their own. Based on this limitation, much research has been done and two additional approaches to using monoclonal antibodies as cancer therapies have emerged. First, it has been found that when monoclonal antibodies are administered in combination with chemotherapy, the antitumor activity is greater than when either therapy is administered alone. Second, monoclonal antibodies that are linked to cell-killing payloads such as drugs or toxins can more effectively kill cancer cells than monoclonal antibodies alone.

Generally speaking, there are four basic methods for using monoclonal antibodies as cancer therapeutics. Each of the approaches described below capitalizes on a monoclonal antibody's ability to precisely target selected molecules that are displayed in high density on the surface of cancer cells:

•

Blocking Cell Activity —monoclonal antibodies can be produced that bind to specific molecules on the cell surface known as receptors to prevent undesirable cell responses, such as proliferation of cancer cells;

•

Activating Cell Activity —monoclonal antibodies can be produced to bind to specific cell surface receptors in order to activate a desired cellular response which may include induction of cell death;

•

Delivering Therapeutic Agents —monoclonal antibodies can be used to deliver cell-killing payloads, such as drugs, radioactive isotopes and toxins, specifically to cancer cells and tissues while minimizing effects on normal cells; and

•

Delivering Enzymes —monoclonal antibodies can be used to deliver enzymes to the surface of cancer cells that are able to convert nontoxic forms of anti-cancer drugs, or prodrugs, into cell-killing drugs within tumor tissue.

27

Approved Monoclonal Antibody Cancer Therapeutics

At the end of 1997, the first monoclonal antibody for cancer therapy, Rituxan, was approved for the treatment of patients with relapsed or refractory, low-grade non-Hodgkin's lymphoma. Rituxan is genetically engineered as a chimeric monoclonal antibody that binds to the CD20 receptor and is jointly marketed in North America by Genentech and IDEC Pharmaceuticals. Worldwide sales of Rituxan were $279 million in 1999 and $444 million in 2000. Rituxan is also being evaluated for treatment in combination with chemotherapy.

Late in 1998, Herceptin, the second monoclonal antibody for cancer therapy, was approved. Usable in approximately 25-30% of patients with breast cancer, Herceptin, engineered as a humanized monoclonal antibody and marketed by Genentech, binds to the HER2 receptor. Although Herceptin is active as a single agent, it is more effective in combination with chemotherapy agents such as Taxol where responses are considerably better than could be obtained with either agent alone. Worldwide sales of Herceptin were $188 million in 1999 and $276 million in 2000.

Early in 2000, the antibody-drug conjugate Mylotarg, the third monoclonal antibody for cancer therapy, was approved for the treatment of relapsed acute myeloid leukemia. Unlike Rituxan and Herceptin that directly attack and kill cancer cells, Mylotarg binds to cell surface receptor identified as CD33 on cancer cells and delivers a drug payload that enters into and kills the cells. Mylotarg is marketed by the Wyeth-Ayerst division of American Home Products.

Our Monoclonal Antibody Technologies

We focus on developing monoclonal antibody-based therapeutics for the treatment of patients with cancer. Four distinct but related technologies form our core business and provide for the discovery and development of an array of unique monoclonal antibody-based anti-cancer therapeutics. Most monoclonal antibodies have the ability to bind to distinct molecules found on the surface of cells, making them desirable as drugs to treat cancer. However, most monoclonal antibodies are not potent enough as single agents to effectively treat cancer. Our technologies enable us to increase the potency, and the efficacy, of monoclonal antibodies with specificity for cancer or related diseases that are not potent enough on their own. These four technologies are:

•

monoclonal antibodies;



•

monoclonal antibodies chemically linked to cell-killing drugs, or monoclonal antibody-drug conjugates;



•

single proteins containing monoclonal antibody and toxin components, or single-chain immunotoxins; and



•

antibody-directed enzyme prodrug therapy, or ADEPT.

28

Monoclonal Antibodies.   Monoclonal antibodies are generally made in mouse form. Our monoclonal antibodies have been genetically modified to reduce or remove their non-human sequences thereby lowering immune response and extending the potential for chronic use in therapy. These monoclonal antibodies can be effective in treating both hematologic malignancies and solid tumors on their own or in combination with chemotherapy. They have lower toxicity than chemotherapy and allow for multiple doses or cycles of therapy. We have several monoclonal antibody candidates in development for treating patients suffering from hematologic malignancies, including SGN-14 targeted to the CD40 receptor and SGN-30 targeted to the CD30 receptor, both of which are currently in preclinical development.

Monoclonal Antibody-Drug Conjugates.   Monoclonal antibody-drug conjugates are composed of monoclonal antibodies that are linked to potent cell-killing drugs. We generally select monoclonal antibodies that bind to receptors that cause the conjugates to enter cells, or internalize. The environment inside the cell causes the cell-killing drug to be released from the monoclonal antibody, allowing it to have the desired cell killing effect. Until released in the cell, the cell-killing drug is inactive, thereby sparing normal cells. In evaluating candidates for our monoclonal antibody-drug conjugate program, we look for chimeric, humanized or human monoclonal antibodies that bind strongly to and enter cancer cells and not most normal cells. An important component of these monoclonal antibody-drug conjugates are the linkers that hold and then release the drugs from the monoclonal antibodies. We have a variety of stable linkers and highly potent cell-killing drugs that can be used with our own, as well as other companies' monoclonal antibodies. Our lead monoclonal antibody-drug conjugate, SGN-15, uses our linker technology to attach the cell-killing drug doxorubicin to the BR96 monoclonal antibody. The BR96 monoclonal antibody binds to a carbohydrate that is found in high density on many different cancer cells including breast, lung, colon, prostate, and ovarian. SGN-15 is currently in three phase II clinical trials in combination with Taxotere to treat patients with breast, colon or prostate cancer.

Single-Chain Immunotoxins.   Our single-chain immunotoxins are comprised of the receptor binding portions of monoclonal antibodies that internalize, combined with toxin components and genetically assembled into single proteins that kill cells by blocking protein production. These single-chain immunotoxins are specific for solid tumors and hematologic malignancies. In addition, we have a novel ribosome-inactivating protein, Bryodin 1, that is only toxic upon entry to the inside of cells. SGN-10, our leading single-chain immunotoxin, is comprised of BR96 and a protein toxin called Pseudomonas exotoxin A. SGN-10 is being evaluated in two phase I clinical trials in patients with breast, lung, colon, prostate, or ovarian cancers.

29

Antibody-Directed Enzyme Prodrug Therapy , or ADEPT .   ADEPT represents a novel approach to minimize drug exposure to normal tissues by using a monoclonal antibody fused to an enzyme. This approach involves the combination of two non-toxic agents to achieve potent antitumor activity while sparing normal tissue. With ADEPT technology, we select monoclonal antibodies that do not internalize but remain bound to the cell surface, distinguishing this technology from our monoclonal antibody-drug conjugate or single-chain immunotoxin technologies. In the first step, a protein containing a monoclonal antibody and enzyme is administered that accumulates on solid tumor masses. This protein converts subsequently administered inactive forms of anti-cancer drugs into potent cell-killing drugs that can penetrate into tumor masses and induce antitumor responses. These effects are significantly greater than those achievable by systemic cell-killing drug administration due to the high drug concentrations that may be achieved within the tumor mass, thereby sparing normal tissue from chemotherapeutic damage. Our lead drug candidate, SGN-17/19, is in development for patients with metastatic melanoma. SGN-17/19 is composed of two agents, SGN-17, a protein containing antibody and enzyme components and SGN-19, a prodrug form of the active compound melphalan. We are conducting ongoing research focused on identifying human enzymes that can activate existing or novel forms of anti-cancer drugs.

Our four technologies provide us with the ability to develop monoclonal antibody-based drug candidates that show antitumor activity alone and direct drugs or toxins to the inside of tumor cells while minimizing the effects on normal tissue. In addition, our technologies allow us to deliver enzymes to tumor cell surfaces that can activate non-toxic forms of inactive anti-cancer drugs into active drugs. These technologies allow us to rapidly convert a variety of our own monoclonal antibodies, as well as monoclonal antibodies from third parties, into drug candidates.

Our Strategy

Our objective is to use our expertise in monoclonal antibodies and our novel technologies to develop our product pipeline and discover new product candidates for the treatment of cancer and related diseases. Our strategy includes initiatives to:

•  Continue to Identify and Develop Novel Monoclonal Antibodies.  In the post-genomic world, thousands of potential new targets are being discovered. Monoclonal antibodies that bind to these targets can be generated rapidly. We believe that monoclonal antibodies will be one of the primary areas for therapeutic development for the foreseeable future, particularly as genomic research identifies new disease targets. We have focused on the research and development of monoclonal antibodies since our inception and have successfully identified and obtained patent rights for several novel monoclonal antibodies with potential therapeutic applications. We are collaborating with Medarex to produce novel fully human monoclonal antibodies to certain breast cancer and melanoma targets. We will continue to apply our expertise in monoclonal antibodies to identify novel monoclonal antibodies that bind to these new targets.

•  Use Our Technologies to Increase Potency of Monoclonal Antibody Therapeutics.  Monoclonal antibodies make excellent delivery vehicles since they bind specifically to cell surface targets. Our expertise and intellectual property rights can be used to make these highly specific monoclonal antibodies into drug candidates by improving the potency and efficacy of monoclonal antibody-based therapeutics through our monoclonal antibody-drug conjugates and antibody-directed enzyme prodrug therapy, or ADEPT, programs. We are also actively developing additional technologies in these programs for which we plan to file patent applications. Our technology provides us with an opportunity to develop our own product candidates, but also enables us to add significant value to monoclonal antibodies and targets owned by other companies.

•  Develop a Broad Portfolio of Products.  We are developing multiple products for many potential indications simultaneously, thereby increasing our opportunities to identify successful drugs. Our drug candidates utilize multiple mechanisms of action and target a variety of different receptors expressed in several types of cancer cells.

30

•  Acquire Attractive Drug Candidates.  In addition to our own development efforts, we will continue to identify products and technologies to in-license. We believe that we are well positioned to continue to attract in-licensing and acquisition candidates as a result of our demonstrated expertise in monoclonal antibodies. We have successfully in-licensed lead monoclonal antibodies from academic groups as well as from other companies. While we expect that many new product candidates will arise from our internal research programs, we will continue to seek in-licensing opportunities to build our product candidate pipeline.

•  Establish Strategic Collaborations.  We intend to enter into corporate collaborations at various stages in the research and development process. We may seek a corporate collaborator prior to initiating phase II clinical trials or may choose to partner some products at a later stage in order to increase our potential downstream participation in product sales. We believe our collaboration strategy provides us with distinct advantages, including:

•

it builds on our fundamental strength in discovery and development of innovative monoclonal antibody-based products and technologies;



•

it capitalizes on our corporate partners' strengths in product development, manufacturing and commercialization;



•

it enables us to develop a greater number of lead agents and programs than otherwise would be possible; and



•

it reduces our financing requirements.

31

Our Clinical and Preclinical Development Programs

We are developing monoclonal antibody-based therapeutics for the treatment of cancer patients. Our research is focused on identifying and characterizing monoclonal antibodies that bind to tumors. Our lead monoclonal antibodies are genetically modified to cause minimal adverse immune system responses. To augment the anti-tumor function of monoclonal antibodies, we link drugs, toxins or enzymes to monoclonal antibodies using our proprietary technologies. After testing our drug candidates in tumor cell culture systems, they are tested in appropriate animal models of human tumors and then clinically in cancer patients.

We currently have two product candidates in clinical development, SGN-15 in phase II trials and SGN-10 in phase I trials, and five product candidates in preclinical development, SGN-14, SGN-30, SGN 17/19, novel BR96 monoclonal antibody-drug conjugate and novel SGN-30 monoclonal antibody-drug conjugate. We are collaborating with Genentech for the development and commercialization of SGN-14. We are also actively engaged in research and discovery of new monoclonal antibodies, targets, linker systems, high-potency drugs and enzymes that can be incorporated into our development portfolio.

The following table summarizes the status of our product candidates currently in clinical trials:

*

Aventis, the manufacturer of Taxotere, is co-funding the clinical trials but has no rights to SGN-15 or SGN-10.

In addition, we have the following product candidates currently in preclinical and discovery stages:

32

Our Product Candidates

SGN-15

SGN-15 is our monoclonal antibody-based drug candidate for treating breast, colon, prostate, lung and ovarian cancers. SGN-15 is a monoclonal antibody-drug conjugate composed of a monoclonal antibody called BR96, genetically modified as a chimeric antibody and chemically linked to the cell-killing drug doxorubicin. BR96 binds to a Lewis ^y -related carbohydrate molecule that is expressed at high levels on many cancer cells, including those of the breast, lung, pancreas, ovary and prostate and on some normal cells in the gastrointestinal tract. SGN-15 works by binding to and entering the cell and then releasing its payload of doxorubicin inside the cell. This release is caused by an acidic environment inside the cell that does not exist outside the cell. Normally, doxorubicin is injected into the body and allowed to circulate throughout the body thereby affecting both cancer and normal tissues. In contrast, our method of targeted drug delivery to the inside of a cell allows for relative sparing of normal tissues from the adverse effects of doxorubicin.

Development Status and Clinical Data.   SGN-15 has been tested as a single agent in three phase I trials and two phase II trials. Based on data from these initial phase I and phase II trials that included 153 patients, SGN-15, as a single agent, localized to human tumors and has antitumor activity. However, the infrequency of this antitumor activity did not support further development as a single agent. Rather, our strategy with SGN-15 is to utilize it in combination with a chemotherapeutic agent.

In September 1999, we initiated a phase I/II trial to study SGN-15 in combination with Taxotere in patients with breast or colon cancer. The rationale for the trial is based on the outstanding antitumor activity of the taxane family of cell-killing drugs, their unique mechanism of action compared to SGN-15, pre-clinical data showing enhanced anti-tumor efficacy of SGN-15 and taxanes, and their non-overlapping toxicity profiles. In our breast cancer trial, we are enrolling patients that have already failed previous chemotherapy, which possibly may have included taxane therapy. In our colon cancer trial, we are enrolling patients that have failed front-line therapy, or the best available approved therapy, and have little or no alternatives for treatment. In September 2000, we completed the phase I component of the phase I/II SGN-15 trial and established a well-tolerated dose of SGN-15 in combination with Taxotere. We safely treated 16 patients and antitumor responses were observed. We initiated separate phase II trials in breast and colon cancer in October 2000 with 14 patients enrolled in the trial as of December 31, 2000. This trial is presently accruing patients at the University of Alabama, Birmingham Cancer Center, Georgetown University Medical Center in Washington, D.C., and the Georgia Cancer Specialists in Atlanta, Georgia. We plan to accrue a total of 30 patients in our phase II colon cancer trial and 45 patients in our phase II breast cancer trial. In order to achieve rapid marketing approval, our phase III development strategy is focused on designing trials for second-line therapy, or those therapies that are available after the front-line therapies have failed.

We recently initiated a third phase II trial in patients with hormone-refractory prostate cancer in combination with Taxotere, a commonly used chemotherapy for this disease. We have observed a synergistic antitumor effect in testing with SGN-15 and taxanes in preclinical prostate cancer models. Patients entering the trial will be placed into two groups of 100 patients each, those treated with the combination of SGN-15 and Taxotere and those treated with Taxotere alone. We plan to include as many as 15 sites in the U.S., with the Arizona Cancer Center being the lead site. The primary endpoints for the trial are a decrease in tumor size, or an objective antitumor response, a measurement of prostate serum antigen and a quality of life assessment. In phase III, we plan to compare SGN-15 and Taxotere to the standard of care for hormone refractory prostate cancer.

In 2001, we plan to initiate two additional phase II trials using the combination of SGN-15 and Taxotere. In the first trial, we plan to enroll patients with non-small cell lung cancer, which represents approximately 80% of all lung cancers. Response rates from approved front-line therapies in these patients are modest and no therapy is curative. In this trial, we plan to treat approximately 40 patients with the combination of SGN-15 and Taxotere and compare the data to approximately 20 patients that are treated with Taxotere alone. Taxotere is commonly used as a second-line therapy for lung cancer with a response rate of less than

33

10 percent. Our second trial is focused on patients with ovarian cancer. Taxotere is one of several cell-killing drugs used as part of front-line and second-line therapeutic strategies for treating ovarian cancer. We plan to test the combination of SGN-15 and Taxotere in second-line treatment and in newly diagnosed patients, who have the most advanced stage of ovarian cancer.

SGN-10

SGN-10 is our single-chain immunotoxin in development for treating breast, colon, lung, pancreatic, ovarian and prostate cancers. Although therapies exist for all of these diseases, no curative treatment exists when the disease is in an advanced stage. SGN-10 is engineered to redirect the potent cell killing activity of a protein toxin called Pseudomonas exotoxin A from its normal target to cancer cells by genetically deleting its natural binding ability and replacing it with the binding capability of our cancer-targeted BR96 monoclonal antibody. BR96 binds to a Lewis ^y -related carbohydrate molecule that is expressed at high levels on many cancer cells, including those of the breast, lung, colon, pancreas, ovary and prostate and on some normal cells in the gastrointestinal tract.

Development Status and Clinical Data.   We are currently conducting a single-agent phase I trial in patients with advanced stage solid tumors. The trial is being conducted at the University of Alabama at Birmingham, the Fox Chase Cancer Center in Philadelphia, Pennsylvania, and the University of Chicago Cancer Center. As of December 31, 2000, a total of 52 patients have been enrolled in the trial. Our development strategy for SGN-10 as a single-agent is to identify appropriate disease targets and conduct disease specific phase II trials. Since we have observed that the majority of patients receiving SGN-10 develop an immune response three weeks after treatment has begun, thus limiting the number of effective doses they can receive, we are investigating strategies to limit the adverse immune response towards SGN-10. These include pre-treating patients with agents that suppress the immune system prior to treating with SGN-10. We plan to test those agents that may prevent a specific immune response without inducing global immune suppression.

In July 2000, we initiated a second phase I trial of SGN-10 in combination with Taxotere to determine the optimal combination dose in patients with advanced stage solid tumors. Our strategy for this trial is to identify a safe combination dose of SGN-10 and Taxotere to utilize in phase II and other advanced trials. Currently, our trials are being conducted at Georgetown University Medical School in Washington, D.C., with other sites to be added at a later time.

Our Preclinical Development Program

We have two separate stages of preclinical development; the first stage involves extensive testing of drug candidates and the second stage involves preparation for clinical development. The initial preclinical stage is focused on the production and testing of drug candidates that we have identified in discovery. We generate material for preclinical testing using our small-scale production and purification capabilities. We then subject our drug candidates to a series of experimental tests that include tumor cell culture, antitumor efficacy in animal models of human cancer and safety trials in animals. Based on our knowledge of the mechanism of drug action and the interactions between drugs, we can test whether two drugs work well together or whether they inhibit the activity of each other. We are able to measure whether two drugs can synergize, or if their combined effect is greater than the sum of their individual effects. We can then compare new drug candidates versus existing therapies in animal models of many different cancer types, including solid tumors, melanomas and hematologic malignancies. From the collective information gathered from these preclinical trials, we can determine the effectiveness and safety of each drug candidate in animal models prior to considering trials in humans.

Our second preclinical development stage is focused on preparing drug candidates for clinical trials, including scale-up production for clinical-grade material and eventual commercialization. Our approach is to develop procedures for large-scale manufacturing and quality control of drug candidates prior to outsourcing

34

or partnering. These standards are compiled in a technology transfer package provided to the selected contract FDA Good Manufacturing Practice manufacturer or corporate partner.

SGN-14 and Related Products

We are collaborating with Genentech to develop a family of anti-cancer agents that target CD40. CD40 is a cell surface receptor that is expressed on a variety of hematologic malignancies such as multiple myeloma, non-Hodgkin's lymphoma and leukemias, certain solid tumors, most notably non-small cell lung, ovarian and bladder cancer, and Kaposi's sarcoma. Our lead anti-CD40 agent is humanized monoclonal antibody SGN-14. Three single-chain immunotoxins, SGN-11, 12 and 18, and an antibody-drug conjugate, SGN-20, are also included in our alliance with Genentech.

SGN-30 and Novel SGN-30 Drug Conjugate

We are developing SGN-30, a monoclonal antibody that targets CD30, for the treatment of hematologic malignancies. CD30 is a cell surface receptor expressed on a variety of hematologic malignancies including Hodgkin's disease, certain leukemias and lymphomas, and in certain immunologic disease indications. We have or are generating chimeric and humanized forms of SGN-30. SGN-30 induces direct anti-cancer activity as a monoclonal antibody. Preclinical trials show that SGN-30 has potent antitumor activity in animal models of human hematologic disease. We are also evaluating high potency monoclonal antibody-drug conjugate forms of SGN-30 in preclinical models. The additional potency of monoclonal antibody-drug conjugates may be useful in treating patients with the most advanced stages of these diseases. The therapeutic utility of SGN-30 and monoclonal antibody-drug conjugate forms of SGN-30 are also being evaluated for use in treating immunologic diseases including multiple sclerosis and lupus.

SGN 17/19

SGN 17/19 is based on our ADEPT technology and is being developed for the treatment of melanoma. ADEPT is an approach to cancer therapy that involves the combination of two non-toxic agents to achieve potent antitumor activity. SGN-17 is a protein containing monoclonal antibody and enzyme components that incorporates the binding site of the monoclonal antibody, L49, and a specific form of the enzyme ß-lactamase. L49 is known to bind to the p97 cell surface molecule, which is non-internalizing and associated with melanoma. p97 is also expressed on many ovarian, breast, and lung carcinomas, although at a lower level. The prodrug, SGN-19, is a form of the chemotherapeutic drug melphalan that has been inactivated through the addition of a chemical group that can be removed by the enzyme ß-lactamase. When SGN-17 is injected systemically, it accumulates on the tumor tissue and remains bound at the cell surface. Once SGN-17 has cleared from the circulation, SGN-19 is then administered systemically. SGN-19 is then converted to melphalan in the tumor tissue by the enzyme ß-lactamase bound to the surface of cancer cells, resulting in localized release of melphalan.

Novel BR96 Monoclonal Antibody Drug Conjugate

In preclinical tests, we have demonstrated enhanced efficacy and lower dose requirements for BR96 monoclonal antibody-drug conjugates by improving the linker chemistry, which connects cell-killing drugs to monoclonal antibodies, and by using proprietary or patented drugs that are more potent than doxorubicin. Thus, as part of our monoclonal antibody-drug conjugate program, we are working to identify second generation, high potency BR96 monoclonal antibody-drug conjugates that have superior characteristics to that of SGN-15. Through a collaboration with Applied Molecular Evolution, we have also identified mutations in the antibody binding site that result in increased binding affinity and an increase in the number of tumors that are recognized. These mutations have been engineered into a humanized version of the antibody. We are collaborating with Genzyme Transgenics, Inc., to produce monoclonal antibodies in the milk of goats. We believe that large quantities of the recombinant protein could be produced at a reasonable cost using this approach.

35

Our Discovery Programs

We have discovery research programs directed towards identifying and developing new monoclonal antibody-based products and technologies to treat cancer. Our discovery programs are currently focused on identifying novel targets, monoclonal antibodies, monoclonal antibody-drug conjugates, monoclonal antibody-drug conjugate technologies and antibody-directed enzyme prodrug therapies.

New Targets and Monoclonal Antibodies

New Targets.   We utilize a variety of genomic tools and biologic assays to identify novel human targets for which we can generate specific new monoclonal antibodies. In the post-genomic era in which vast amounts of DNA sequence information is available, the opportunity to identify novel target genes is more accessible than ever before. We focus on genes that are highly expressed in cancer and virally-infected tissue to identify molecules that are located on the surface of cancer cells that may serve as targets for monoclonal antibodies. This may be in the form of monoclonal antibody-drug conjugates as well as single proteins containing monoclonal antibody and enzyme components that can provide for selective anti-cancer prodrug activation within the tumor. In addition to internal discovery efforts, we have an active in-licensing program focused on novel cancer targets to which we generate new monoclonal antibodies.

New Monoclonal Antibodies.   We are collaborating with Medarex to use certain breast cancer and melanoma targets to generate novel fully human monoclonal antibodies. These monoclonal antibodies may represent product candidates on their own or may be suited to kill cancer cells by utilizing our technologies such as monoclonal antibody-drug conjugates. We have also established a process whereby in-licensed monoclonal antibodies can be developed into drug candidates. For example, we in-licensed SGN-14 and SGN-30 as

36

mouse monoclonal antibodies, and then genetically modified these monoclonal antibodies to be used alone and/or with our payload technologies. SGN-14 and SGN-30 have been or are in the process of being genetically modified in chimeric and humanized forms. At present, we are evaluating additional opportunities with academic groups and other companies to in-license mouse and genetically modified monoclonal antibodies in chimeric, humanized, or fully human forms.

Additionally, we have monoclonal antibodies that bind to cancer cells and are being evaluated as therapeutic agents. These include the monoclonal antibody BR110 which binds strongly to carcinomas of the lung, colon, breast, and ovary. BR110 binds to the GA-733-1 molecule, a related family member to the target of Panorex, which has shown efficacy in patients with colon carcinoma. Panorex is currently marketed in Germany, and is in phase III trials sponsored by Glaxo in the U.S. BR110 has the ability to internalize cell-killing drug payloads into tumor cells and can be utilized with our monoclonal antibody-drug conjugate technology. We are also using the high-affinity forms of the BR96 monoclonal antibody that react with the Lewis ^y -related and Lewis ^x -related molecules that are expressed on most solid tumor types. Specifically, both new BR96 monoclonal antibodies in humanized form and the chimeric form of BR96 are being linked to high potency cell-killing drugs to form monoclonal antibody-drug conjugates. We have also identified monoclonal antibodies that bind to ovarian cancer cells and are in early stages of evaluation.

Monoclonal Antibody-Drug Conjugate Program

We are engaged in the discovery and development of novel, high potency monoclonal antibody-drug conjugates. The key elements for preparing highly effective drug conjugates are monoclonal antibody specificity, drug potency and linker technology.

Monoclonal Antibody-Specificity.   We use monoclonal antibodies that strongly bind to cancer cells while minimally binding to normal tissues. For monoclonal antibody-drug conjugates, we use monoclonal antibodies that enter the cell after binding to its surface.

Drug Potency.   We have identified two different cell-killing drug types with unique mechanisms of action that are each approximately 1000-fold more potent than doxorubicin, the cell-killing drug component of SGN-15. One of these is a class of drug known as minor groove binders that bind to DNA and inhibit DNA replication, thereby killing cells. Our minor groove binders can be synthesized and conjugated to monoclonal antibodies. This class of cell-killing drug is not subject to pathways found in tumors that make them resistant to chemotherapeutic drugs. We are presently focused on the design and testing of potent minor groove binders that we believe will provide us with a proprietary position in this area. The monoclonal antibody-drug conjugates formed from minor groove binders are highly active at clinically relevant doses.

Our second drug type is known as an anti-mitotic agent since it functions by inhibiting tumor cell division. Auristatin E, a highly potent, synthetic derivative of the natural product Dolostatin 10, has been found to have antitumor activity against human melanoma and lung carcinomas implanted in mice. The chemical structure of Auristatin E contains a chemical group making it available for conjugation with monoclonal antibodies. Due to its properties and our ability to synthesize it, we believe Auristatin E represents a promising candidate for targeted delivery to cancer cells and may also be useful by itself as an antitumor or chemotherapy agent.

Linker Technology.   We have obtained rights to a series of proprietary, highly stable, peptide-based linkers with issued worldwide patents. Many of these peptide-based linkers are cleaved by specific enzymes called proteases that function inside cells. Importantly, once cleaved, the linkers are self-degrading so that the released drug is chemically unmodified and has full potency. Monoclonal antibody-drug conjugates using our linkers have been found to be highly effective antitumor agents with long-term stability in human plasma. We have developed several other linker technologies through internal research efforts. Our new linkers include those that allow drugs to be released when they enter the acidic environment within tumor cells, and by several different enzymes within cells. Currently, linkers are produced separately from the cell-

37

killing drug and monoclonal antibody components of monoclonal antibody-drug conjugates. To simplify production, we have developed linkers that are incorporated directly into the cell-killing drug component.

Antibody-Directed Enzyme Prodrug Therapy; ADEPT

ADEPT is an approach to cancer therapy that involves the combination of two non-cell-killing agents to achieve potent antitumor activity. First, we administer a single protein containing monoclonal antibody and enzyme components that can localize within solid tumor masses. These proteins remain bound to cell surface molecules, and activate subsequently administered inactive anti-cancer drug derivatives. Upon activation, potent drugs are released that can penetrate into tumor masses and induce antitumor responses. The effects are significantly greater than those achievable by systemic cancer drug administration due to the high drug concentrations reached inside the tumor.

We have identified several enzyme and prodrug combinations that lead to high levels of antitumor activity at well tolerated doses. These include recombinant proteins containing monoclonal antibody and enzyme components that activate prodrug forms of existing drugs, including melphalan, mitomycin C, paclitaxel and doxorubicin. We have an ongoing effort to develop human enzymes for prodrug activation, and have identified enzymes capable of activating the clinically approved drug CPT-11, which is used for the treatment of advanced colon cancer.

Single-Chain Immunotoxins

We have prepared a variety of single-chain immunotoxins targeted to molecules expressed on the surface of cancer cells. We have obtained rights to Bryodin 1, a patented ribosome-inactivating protein that is a preferred component compared to toxins currently used in the construction of single-chain immunotoxins. Bryodin 1 is especially useful due to its potency to kill once inside cells and its lack of toxicity to animals.

Corporate Collaborations

Part of our strategy is to establish corporate collaborations with pharmaceutical, biopharmaceutical and diagnostic companies. We plan to collaborate with others, both for the development and commercialization of our own drug candidates and for the potential improvement of collaborators' monoclonal antibodies using our technologies. We focus our efforts on partnering our technologies at various stages in the research and development process. We target collaborators that have the expertise and capability to develop, manufacture, obtain regulatory approval for and commercialize our monoclonal antibody-based products. In our corporate collaborations, we seek to cover our research and deve1opment expenses through research funding, milestone payments and option, technology or license fees. We also seek to retain significant downstream participation in product sales through either profit-sharing or product royalties paid on annual net sales.

Genentech

In June 1999, we licensed our CD40 agents, including SGN-14, SGN-18 and SGN-20, on an exclusive basis to Genentech and granted Genentech an option under specific circumstances to license SGN-11 and SGN-12. SGN-14 is currently in preclinical development for the treatment of patients with hematologic malignancies or other types of cancer. Our agreement with Genentech includes joint oversight of development. However, costs and tasks may not be assigned to us without the approval of our representative on the joint oversight committee. We do not anticipate that these costs or tasks will be assigned to us in the future. The business terms of this agreement include potential royalties on net sales as well as other payments of up to $45.0 million, including $41.0 million in potential milestone payments on the first product developed. The agreement also provides for milestone payments of up to $20.0 million and future royalties on net sales of each additional product incorporating our technology. Genentech's obligation to pay us royalties under this agreement terminates on a country-by-country and product-by-product basis upon the later of a specified number of years after the first commercial sale in each country or the last to expire of

38

the licensed patents in each country. The agreement is also subject to earlier termination by Genentech at any time upon 90 days notice or by either party if the other party enters bankruptcy or breaches its material obligations thereunder.

We believe that partnering with Genentech will optimize time to market by utilizing their existing development, manufacturing capabilities, marketing and sales force. As part of this agreement, we sold Genentech 680,272 shares of Series B convertible preferred stock in December 1999 and will sell 166,667 registered shares of common stock to Genentech directly in this offering at the initial public offering price.

Bristol-Myers Squibb

We obtained the rights to some of our technologies and drug candidates through a license agreement with Bristol-Myers Squibb, portions of which are exclusive. Through this license, we secured rights to Bristol- Myers' monoclonal antibody-based cancer targeting program, which includes rights to 24 different patents, eight monoclonal antibodies, chemical linkers, a ribosome-inactivating protein and enabling technologies. Our license encompasses four technologies: genetically modified monoclonal antibodies, monoclonal antibody-drug conjugates, single-chain immunotoxins and antibody-directed enzyme prodrug therapies. Under this license agreement, we received FDA Good Manufacturing Practices produced and vialed material for two different monoclonal antibody-based therapeutic agents, SGN-15 and SGN-10, which have entered clinical trials. Under the terms of the license agreement, we are required to pay royalties on net sales of future products incorporating the licensed technology. Our obligation to pay royalties under this agreement terminates on a product-by-product and country-by-country basis upon the later of ten years after the first commercial sale in each country or the last to expire of the licensed patents in each country. The last of the licensed issued patents will expire in 2018, although our obligation to pay royalties to Bristol-Myers may continue for a longer period of time to the extent that we use technology derived from pending patent applications or other subsequently issued patents. The agreement is also subject to earlier termination if either of the parties breaches its material obligations thereunder.

A portion of the technology we have licensed from Bristol-Myers is derived through sublicenses of technology from the following third parties:

Enzon.   We have sublicensed rights to single-chain antigen binding molecules from Enzon, Inc. This technology is used in SGN-10. We have semi-exclusive rights to the Enzon technology, subject to the rights of several existing Enzon licensees, as well as our obligation to make biannual payments to maintain this semi-exclusivity through 2003. Under the terms of our sublicense with Enzon, we are also required to make milestone payments and pay royalties on net sales of products incorporating technology sublicensed from Enzon. Our obligation to pay royalties under this agreement terminates on a country-by-country basis upon the last to expire of the licensed patents in each country. The last of the licensed issued patents will expire in 2007, although our obligation to pay royalties to Enzon may continue for a longer period of time to the extent that we use technology derived from pending patent applications or other subsequently issued patents. The agreement is also subject to earlier termination if either party breaches its material obligations thereunder.

Applied Molecular Evolution.   We have sublicensed exclusive rights to certain humanized forms of the BR96 monoclonal antibody from Applied Molecular Evolution. Under the terms of our sublicense, we are required to make milestone payments and royalties on net sales of products incorporating technology sublicensed from Applied Molecular Evolution. Our obligation to pay royalties under this agreement terminates on a product-by-product and country-by-country basis upon the later of ten years after the first commercial sale in each country or the last to expire of the licensed patents in each country. The last of the licensed issued patents will expire in 2015, although our obligation to pay royalties to Applied Molecular Evolution may continue for a longer period of time to the extent that we use technology derived from pending patent applications or other subsequently issued patents. The agreement is also subject to earlier termination if either of the parties breaches its material obligations thereunder.

39

Medarex.   We are collaborating with Medarex to produce fully human monoclonal antibodies to certain breast cancer and melanoma antigen targets identified by us over the next three years in order to develop and commercialize monoclonal antibody-based products. The agreement calls for joint development of at least half of our breast cancer antigens and a specific melanoma antigen. There will be a joint steering committee composed of members of both companies to make development decisions concerning jointly developed monoclonal antibody product candidates. Under the agreement, all development, manufacturing, and clinical costs of jointly developed products and all net profits or net losses will be shared by us and Medarex. Each of us has the right to opt out of the joint development of any antigen target and receive instead certain milestone and royalty payments on net sales. The agreement terminates upon the later of one year after completion of the research activities thereunder or the date on which neither party is exploiting any jointly developed products. The agreement is also subject to termination if either party enters bankruptcy or breaches its material obligations thereunder. As part of this agreement, we will sell to Medarex $2.0 million of our common stock at the initial public offering price in a private placement concurrent with this offering.

We have also licensed and intend to continue to license product and marketing rights from selected commercial, research and academic institutions in order to capitalize on the capabilities and technology bases of these entities, including the following:

ICOS Corporation.   In October 2000, we entered into a license agreement with ICOS Corporation for non-exclusive rights to use the CHEF expression system, a DNA sequence we may use to manufacture SGN-30. Under the terms of our agreement with ICOS, we are required to make milestone payments and pay royalties on net sales of products manufactured using the CHEF expression system. Our obligation to pay royalties under this agreement terminates upon the expiration of the last to expire of the licensed patents, which will occur in 2017, although our obligation to pay royalties to ICOS may continue for a longer period of time to the extent that we use technology derived from pending patent applications or other subsequently issued patents. The agreement is also subject to earlier termination if either party enters bankruptcy or defaults in the performance of any material provision thereunder.

Mabtech AB.   In June 1998, we obtained exclusive worldwide rights to a monoclonal antibody that recognizes CD40 from Mabtech AB. Under the terms of our license with Mabtech, we are required to make a milestone payment and pay royalties on net sales of products incorporating technology licensed from Mabtech. Our obligation to pay royalties under this agreement terminates ten years after the first commercial sale of a product incorporating Mabtech's technology. The agreement is also subject to earlier termination if either party breaches its material obligations thereunder.

University of Miami.   In September 1999, we entered into an exclusive license agreement with the University of Miami, Florida, covering an anti-CD30 monoclonal antibody that is the basis for two new drug candidates targeted to hematologic malignancies and immune system disease. Under the terms of our license with the University of Miami, we made an up-front payment and are required to make milestone payments, annual maintenance fee payments and pay royalties on net sales of products incorporating technology licensed from the University of Miami. Our obligation to pay royalties under this agreement terminates ten years after the first commercial sale of a product incorporating the University of Miami's technology. The agreement is also subject to earlier termination by the University of Miami if we enter bankruptcy or by either party if the other party breaches its material obligations thereunder.

Arizona State University.   In February 2000, we entered into a license agreement with the Arizona State University covering the cell-killing agent Auristatin E, a synthetic derivative of the natural product Dolastatin 10. We intend to use Auristatin E as a component of new monoclonal antibody-drug conjugates. We are also testing Auristatin E in preclinical models to determine whether it qualifies for single-agent clinical development. Under the terms of our license with Arizona State University, we are required to make milestone payments, annual maintenance fee payments and pay royalties on net sales of products incorporating technology sublicensed from Arizona State University. Our obligation to pay royalties under

40

this agreement terminates on a country-by-country basis upon the last to expire of the licensed patents in each country, which will occur in 2014, although our obligation to pay royalties to Arizona State University may continue for a longer period of time to the extent that we use technology derived from pending patent applications or other subsequently issued patents. The agreement is also subject to earlier termination if either of the parties enters bankruptcy or breaches its material obligations thereunder.

Aventis Pharmaceuticals.   Aventis co-funds three different clinical trials using two of our drug candidates, SGN-15 and SGN-10. In the SGN-15 program, which is being tested in combination with the widely used chemotherapeutic agent Taxotere, Aventis is funding 50% of the clinical trial costs directly to the clinical sites and providing Taxotere drug product. The SGN-15 trials that are part of the Aventis co-funding agreement include a phase I/II trial in breast or colon cancer patients for which the phase I was completed in September 2000 and two separate phase II trials in breast and colon cancers initiated in October 2000. As part of the SGN-10 program, Aventis is co-funding a phase I trial in patients including those with breast, colon, lung, prostate, ovarian or pancreatic cancer. In this setting, SGN-10 is being tested in combination with Taxotere to determine the appropriate dose and disease indication for later stage clinical testing. Aventis is funding 50% of the clinical trial costs directly to the clinical centers and providing Taxotere drug product. Aventis does not obtain any rights or options to SGN-15 or SGN-10 under the co-funding arrangement.

Genzyme Transgenics.   We have agreed to collaborate with Genzyme Transgenics to determine clinical and commercial potential for a form of the humanized monoclonal antibody hBR96-2. Under the terms of the agreement, Genzyme Transgenics will supply us with hBR96-2 and we will perform experiments to evaluate the material. If development is continued by us, we will use good faith efforts to negotiate with Genzyme Transgenics to enter into a supplier agreement for manufacturing of hBR96-2.

Creative Biomolecules.   In September 1998, we entered into a non-exclusive license agreement with Creative Biomolecules, Inc. for the rights to use single-chain antibody technology. We use this technology in SGN-10. Under the terms of this license agreement, we are required to make milestone payments, annual maintenance fee payments and pay royalties on net sales of products incorporating technology licensed from Creative Biomolecules. Our obligation to pay royalties under this agreement terminates on a country-by-country basis upon the last to expire of the licensed patents in each country, which will occur in 2015, although our obligation to pay royalties to Creative Biomolecules may continue for a longer period of time to the extent that we use technology derived from pending patent applications or other subsequently issued patents, subject to extension if we use technology licensed under subsequently issued divisional patents. The agreement is also subject to earlier termination if either party breaches its material obligations thereunder.

Brookhaven Science Associates, LLC.   In January 1998, we entered into a non-exclusive license agreement with Brookhaven Science Associates, operator of Brookhaven National Laboratory, to secure the rights to use the T7 promoter, a DNA sequence we use to manufacture SGN-10. Under the terms of this agreement, we are required to make annual maintenance fee payments and pay royalties on net sales of products manufactured using the T7 promoter. Our obligation to pay royalties under this agreement terminates upon the last to expire of the licensed patents, which will occur in 2017, although our obligation to pay royalties to Brookhaven may continue for a longer period of time to the extent that we use technology derived from pending patent applications or other subsequently issued patents. The agreement is also subject to earlier termination by Brookhaven if we breach our material obligations thereunder and by us at any time upon six months notice.

Public Health Service.   In September 1998, we entered into a non-exclusive license agreement with the Public Health Service for the rights to use truncated forms of a protein toxin called Pseudomonas exotoxin, one of which is a component of SGN-10. Under the terms of this agreement, we are required to make milestone payments and pay royalties on net sales of products incorporating technology licensed from the Public Health Service. Our obligation to pay royalties under this agreement terminates upon the last to

41

expire of the licensed patents, which will occur in 2016, although our obligation to pay royalties to the Public Health Service may continue for a longer period of time to the extent that we use technology derived from pending patent applications or other subsequently issued patents. The agreement is also subject to earlier termination by the Public Health Service if we enter bankruptcy or by either party if the other party breaches its material obligations thereunder.

Genentech; Cabilly License.   In January 2000, we entered into a co-exclusive license with Genentech for rights to use chimeric monoclonal antibodies targeted to Lewis ^y . We are currently using this technology in our product candidate SGN-15. We are performing collaborative research in lieu of an up-front cash payment, and are required to make milestone payments and must make minimum annual royalty payments beginning in 2003 under the terms of the license agreement. Our obligation to pay royalties under this agreement terminates on a country-by-country basis upon the last to expire of the licensed patents in each country, which will occur in 2006, although our obligation to pay royalties to Genentech may continue for a longer period of time to the extent that we use technology derived from pending patent applications or other subsequently issued patents. The agreement is also subject to earlier termination if either party enters bankruptcy. Genentech may terminate the agreement if we breach our material obligations thereunder and we may terminate the agreement at any time.

During the next 24 months, we expect to pay up to an aggregate of approximately $600,000 in annual maintenance fees and milestone payments under all of our license and collaboration agreements combined. We do not expect to pay any royalties on net sales of products under any of these agreements for at least the next several years. The milestone payments could be substantially higher and the royalties could be payable earlier if we file or receive regulatory approvals or achieve commercial sales sooner than expected.

Patents and Proprietary Technology

Our success will depend in large part on our and our licensors' abilities to:

•

obtain patent and other proprietary protection for antigens, antibodies, adjuvants and delivery systems;

•

defend patents once obtained;

•

preserve trade secrets; and

•

operate without infringing the patents and proprietary rights of third parties.

We intend to seek appropriate patent protection for our proprietary technologies by filing patent applications in the United States and certain other countries. As of December 31, 2000, we owned or had licensed 24 issued United States patents, and 7 pending United States patent applications.

These patents and patent applications are directed to certain monoclonal antibodies, drug candidates, linker technologies, monoclonal antibody-drug conjugate technologies, immunotoxin technologies, monoclonal antibody-enzyme and prodrug technologies and enabling technologies. Although we believe our patents and patent applications provide a competitive advantage, the patent positions of pharmaceutical and biopharmaceutical companies are highly uncertain and involve complex legal and factual questions. For example, there is substantial uncertainty regarding the potential for patent protection for gene fragments or genes without known function or correlation with specific diseases. We and our corporate collaborators or licensors may not be able to develop patentable products or processes. We and our corporate collaborators or licensors may not be able to obtain patents from pending patent applications. Even if patent claims are allowed, the claims may not issue, or in the event of issuance, may not be sufficient to protect the technology owned by or licensed to us or our corporate collaborators.

Our or our corporate collaborators' current patents, or patents that issue on pending applications, may be challenged, invalidated, infringed or circumvented, and the rights granted in those patents may not provide proprietary protection or competitive advantages to us.

42

Patent applications in the United States have been maintained in secrecy until patents issue and patent applications in certain foreign countries generally are not published until many months or years after they are filed. Scientific and patent publication often occurs long after the date of the scientific developments disclosed in those publications. Accordingly, we cannot be certain that we or one of our corporate collaborators was the first to invent the subject matter covered by any patent application or that we or one of our corporate collaborators were the first to file a patent application for any such invention.

Our commercial success depends significantly on our ability to operate without infringing patents and proprietary rights of third parties. A number of pharmaceutical companies, biotechnology companies, universities and research institutions may have filed patent applications or may have been granted patents that cover technologies similar to the technologies owned, optioned by or licensed to us or our corporate collaborators. We cannot determine with certainty whether patents or patent applications of other parties may materially affect us or our corporate collaborators' ability to make, use or sell any products.

The existence of third-party patent applications and patents could significantly reduce the coverage of the patents owned, optioned by or licensed to us or our corporate collaborators and limit our or our corporate collaborators' ability to obtain meaningful patent protection. If patents containing competitive or conflicting claims are issued to third parties, we or our corporate collaborators may be enjoined from pursuing research, development or commercialization of products or may be required to obtain licenses, if available, to these patents or to develop or obtain alternative technology. In addition, other parties may duplicate, design around or independently develop similar or alternative technologies to ours, our corporate collaborators or our licensors. If another party controls patents or patent applications covering our products, we and our corporate collaborators may not be able to obtain the rights we need to those patents or patent applications in order to commercialize our products.

Litigation may be necessary to enforce patents issued or licensed to us or our corporate collaborators or to determine the scope or validity of another party's proprietary rights. United States Patent Office interference proceedings may be necessary if we and another party both claim to have invented the same subject matter.

We could incur substantial costs if:

•

litigation is required to defend against patent suits brought by third parties;

•

we participate in patent suits brought against or initiated by our corporate collaborators;

•

we initiate similar suits; or

•

we participate in an interference proceeding.

In addition, we or our corporate collaborators may not prevail in any of these actions or proceedings. An adverse outcome in litigation or an interference or other proceeding in a court or patent office could:

•

subject us to significant liabilities;

•

require disputed rights to be licensed from other parties; or

•

require us or our corporate collaborators to cease using certain technology.

We also rely on trade secrets and proprietary know-how, especially when we do not believe that patent protection is appropriate or can be obtained. Our policy is to require each of our employees, consultants and advisors to execute a confidentiality and inventions agreement before beginning their employment, consulting or advisory relationship with us. These agreements generally provide that the individual must keep confidential and not disclose to other parties any confidential information developed or learned by the individual during the course of their relationship with us except in limited circumstances. These agreements also generally provide that we shall own all inventions conceived by the individual in the course of rendering services to us.

We work with others in our research, development and commercialization activities. Disputes may arise about inventorship and corresponding rights in know-how and inventions resulting from the joint creation or

43

use of intellectual property by us and our corporate partners, licensors, scientific collaborators and consultants. In addition, other parties may circumvent any proprietary protection we do have. As a result, we may not be able to maintain our proprietary position.

Government Regulation

Our products are subject to extensive regulation by numerous governmental authorities, principally the FDA, as well as numerous state and foreign agencies. We need to obtain clearance of our potential products by the FDA before we can begin marketing the products in the United States. Similar approvals are also required in other countries.

Product development and approval within this regulatory framework is uncertain, can take a number of years and requires the expenditure of substantial resources. The nature and extent of the governmental premarket review process for our potential products will vary, depending on the regulatory categorization of particular products. We believe that the FDA and comparable regulatory bodies in other countries will regulate monoclonal antibody products and related pharmaceutical products as biologics. The necessary steps before a new biological product may be marketed in the United States ordinarily include:

•

preclinical laboratory and animal tests;

•

submission to the FDA of an investigational new drug application which must become effective before clinical trials may commence;

•

completion of adequate and well controlled human clinical trials to establish the safety and efficacy of the proposed drug for its intended use;

•

the submission to the FDA of a biologics license application; and

•

FDA review and approval of the biologics license application prior to any commercial sale or shipment of the product.

Preclinical tests include laboratory evaluation of the product, as well as animal trials to assess the potential safety and efficacy of the product. Preclinical tests must be conducted by laboratories that comply with FDA regulations regarding good laboratory practices. The results of preclinical tests, together with manufacturing information, analytical data and proposed clinical trial protocols, are submitted to the FDA as part of an investigational new drug application, which must become effective before the commencement of clinical trials. The investigational new drug application will automatically become effective 30 days after receipt by the FDA unless the FDA indicates prior to the end of such 30-day period that the proposed protocol raises concerns that must be resolved to the satisfaction of the FDA before the trials may proceed. In such case, we cannot assure you that this resolution will be achieved in a timely fashion, if at all. In addition, the FDA may impose a clinical hold on an ongoing clinical trial if, for example, safety concerns arise, in which case the trial cannot recommence without FDA authorization under terms sanctioned by the agency.

Clinical trials involve the administration of the product to healthy volunteers or to patients under the supervision of a qualified principal investigator. Clinical trials are conducted in accordance with good clinical practices under protocols that detail the objectives of the trial, inclusion and exclusion criteria, the parameters to be used to monitor safety and the efficacy criteria to be evaluated. Each protocol must be submitted to the FDA as part of the investigational new drug application. Further, each clinical trial must be reviewed and approved by an independent institutional review board at the institutions at which the trial will be conducted. The institutional review board will consider, among other things, ethical factors and the safety of human subjects. The institutional review board may require changes in a protocol, and there can be no assurance that the submission of an investigational new drug application will enable a trial to be initiated or completed.

Clinical trials generally are conducted in three sequential phases that may overlap. In phase I, the initial introduction of the product into healthy human subjects or patients, the product is tested to assess safety,

44

metabolism, pharmacokinetics and pharmacological actions associated with increasing doses. Phase II usually involves trials in a limited patient population to:

•

determine the efficacy of the potential product for specific, targeted indications;

•

determine dosage tolerance and optimum dosage; and

•

further identify possible adverse reactions and safety risks.

If a compound is found to be effective and to have an acceptable safety profile in phase II evaluations, phase III trials are undertaken to evaluate further clinical efficacy in comparison to standard therapies, within a broader patient population, generally, at geographically dispersed clinical sites. Phase I, phase II or phase III testing may not be completed successfully within any specific period of time, if at all, with respect to any of our potential products. Furthermore, the FDA or an institutional review board or we may suspend a clinical trial at any time for various reasons, including a finding that the subjects or patients are being exposed to an unacceptable health risk.

The results of pharmaceutical development, preclinical trials and clinical trials are submitted to the FDA in the form of a biologics license application for approval of the manufacture, marketing and commercial shipment of the biological product. The testing and approval process is likely to require substantial time, effort and resources, and there can be no assurance that any approval will be granted on a timely basis, if at all. The FDA may deny a biologics license application if applicable regulatory criteria are not satisfied, require additional testing or information, or require postmarket testing and surveillance to monitor the safety or efficacy of the product. In addition, after marketing approval is granted, the FDA may require post-marketing clinical trials, which typically entail extensive patient monitoring and may result in restricted marketing of an approved product for an extended period of time.

Please see the table in "Our Clinical and Preclinical Programs" on page 32 for a complete description of the status of our FDA applications for our product candidates.