Heat Shock Proteins’ Vaccine Potential: From Basic Science Breakthroughs to Feasible Personalized Medicine

by Bruce Goldman

One night in the early 1960s, somebody in a Drosophila lab in Italy inadvertently cranked up the incubator thermostat a bit too high. The next day, the fruit flies’ salivary gland chromosomes exhibited odd puffing patterns under microscopic examination, indicating an unusual pattern of gene expression. Thus began the exploration of a group of related molecules bearing the name “heat shock proteins” (HSPs, for short). That name vastly understates the HSP family’s astounding versatility.

Whenever a cell — any cell, in any organism — is stressed by heat, cold, or glucose or oxygen deprivation, to name but a few examples, HSPs are induced. In fact, the members of this family of dozens of related proteins account for an astounding 20 percent of a stressed cell’s soluble protein contents.

HSPs abound even under perfectly tranquil conditions, when they comprise on the order of 2 percent of the cell’s contents. For, it’s now known, HSPs are involved not only in nursing proteins back to conformational health when they droop, but in other ‘chaperone’ functions as well: midwifing newborn proteins into the proper conformation to begin with; hustling them around among the cytosolic compartments; and, when they’re irredeemably bent out of shape, carting them off to intracellular garbage disposals to be degraded into the short chains of amino acids called peptides.

Yet another key role suspected of HSPs is loading peptide detritus onto another important class of proteins called major histocompatibility complex, or MHC, molecules. A loaded MHC molecule heads for the cell surface, where it can be monitored by roving sentry cells of the immune system that react vigorously if they sense any peptides that shouldn’t be there — for instance, fragments of viral proteins or of altered constituents of a cancerous cell.

With so many important functions, it’s little wonder that HSPs are phenomenally well conserved across and within species. The similarity between a human’s HSPs and those of a mouse is greater than 95 percent. Even bacterial HSPs bear something like a 50 percent homology to ours.

All the roles just described for HSPs have been intracellular. They are essentially confined to a healthy cell’s cytosol and other intracellular compartments. But nature, ever the economizer, would be remiss if it were to let this kind of talent go to waste.

Library Science

In 1980 — back before HSPs’ intracellular functions, outlined above, had been fully delineated — a biochemistry graduate student in Hyderabad, India, named Pramod Srivastava was musing upon the well established observation that rats injected with attenuated cells from a tumor to which they are susceptible will reject a new graft of that very tumor, whereas rats not thus inoculated will prove lethally hospitable to it. Srivastava began fractionating tumor lysates and injecting these fractions into rodents. He then fractionated the fractions and so forth, until he had isolated the apparent immunizing agents: heat shock proteins.

Although Srivastava published his results in 1984, they received scant academic attention until a decade later, when they were reproduced by an independent group in Heidelberg, Germany. By that time, Srivastava had co-founded a company, New York-based Antigenics, on the principle that HSPs could be used as immunotherapeutic vaccines. One of the things Srivastava had learned was that the immune response HSPs induced was highly specific: HSPs isolated from Tumor X could immunize against Tumor X, but not against Tumor Y, and vice versa.

HSPs’ exquisite specificity in priming animals’ immune systems was odd in light of the fact that HSPs were apparently without allelic variation. That mystery was resolved when Srivastava showed that it was not HSPs per se but the peptides cradled noncovalently in their binding sites that were somehow conveying antigenic intelligence to the immune system. If you dissociate the peptide from the HSP — even if you leave the peptide in the mix — immunogenicity is lost.

Given HSPs’ role as all-purpose cellular chaperones, it’s not surprising that they would excel at grabbing onto peptides, nor that they would do so rather promiscuously. Between them all, HSPs are believed to bind about every conceivable peptide, regardless of size, composition or water- vs. fat-solubility.

The result is that a cell’s entire inventory of HSPs is a ‘library’ containing copies of every peptide ‘book’ the cell has produced lately. Most of those books are ones the immune system read long ago, during its early development, and no longer finds of interest because they are peptides a healthy cell should have. (During their training, immune cells responding too vigorously to ‘self’ antigens are downregulated or dispatched, resulting in a state called immunotolerance to these antigens.) But HSPs inside a virally infected or cancerous cell are also holding onto peptides that are foreign, inappropriately expressed, or altered: In other words, peptides that a healthy cell ought not to have, and that might serve as red flags for the immune system — if only there were a way for the immune system to determine the contents of these rather private libraries.

Fortunately, there is.

Two Kinds of Cell Death

Under anything resembling normal circumstances, HSPs simply don’t show up in extracellular fluids. But cells do die, in one of two ways. Sometimes a cell dies in response to an externally or internally issued suicide command. Such programmed cell death — such as that of a skin cell in the web between an embryonic human’s fingers as the web shrivels away — is called apoptosis, which is orchestrated and tidy. The cell’s contents (including its entire HSP-peptide library) are sequestered in so-called apoptotic bodies, later to be cleaned up through ingestion by antigen-presenting cells (APCs) such as macrophages and dendritic cells. Recent evidence, from Srivastava’s lab as well from molecular biologist Richard Vile at the Mayo Clinic in Rochester, Minn., and others, shows that apoptotic bodies ingested by APCs cause changes on APC surfaces that actually downregulate the immune system, inducing tolerance instead of activation.

For an antigen to stimulate T-cell activation it must be presented on a cell’s surface accompanied not only by an appropriate MHC molecule but also by a 'costimulator' molecule called B-7. Apoptotic bodies, the HSPs of which are opaque to APC inspection, appear to downregulate the expression of B-7 on APC surfaces, leading to T-cell apathy rather than activation.

But cells can also meet an untimely demise. A tumor is a collection of abnormally growing cells. When — due, for example, to the overpowering burden of mutagenesis — a cell undergoes the messy, unplanned death called necrosis, it bursts like a water balloon, spilling its internal contents into the surrounding medium. Among the jettisoned debris bob the HSPs, carrying among them all the peptide suspects an immune surveillance cell could possibly want to see.

Cancer cells, in the course of their rapid division, may evolve ways of suppressing immune activity in their immediate neighborhood (for example, by secreting immunosuppressive chemicals) or of hiding their identifying surface antigens (for example, behind a veil of attached sugar groups). But in their ferment, most tumors toss off a constant effluvium of necrotic cells, exposing their HSPs to the immune system. And when a dendritic cell or macrophage takes up and re-presents HSP-associated peptides, B-7 expression is upregulated, creating the optimal environment for T-cell activation.

Srivastava’s group recently identified a receptor (called CD91) on the surface of dendritic cells for members of the HSP family, further nailing down his once-heretical case for HSP involvement in the immune cascade. HSP receptors have also turned up on macrophages and platelets. Scientists are still arguing over the details of precise molecular mechanisms but data emanating from high-powered labs around the world have fostered a consensus around these major points: HSPs released in necrotic conditions actively draw phagocytic APCs (particularly dendritic cells) into the area, hand over peptides to the incoming APCs, and secrete chemicals, predisposing them to re-present the HSP-delivered antigens in ways that encourage active response by the immune system’s cellular branch. Compared with the B cell-driven humoral, or antibody, response, this cellular process is far better suited to attacking tumors and virally infected cells.

Each of the key elements of the above hypothesis, first advanced in the early 1990s by Srivastava, has since been confirmed by independent labs at Harvard, Rockefeller and Tubingen universities; the Mayo Clinic; Karolinska and Max Planck institutes; and elsewhere. “There’s no ambiguity about it,” says Nobelist and Rockefeller University Professor Emeritus Joshua Lederberg, who met Srivastava several years ago and is now honorary chairman of Antigenics’ scientific advisory board. “Pramod proceeded very methodically and with great insight, and with no preconceptions.”

Yet classical immunology would not have predicted this because cellular immune responses are generally triggered only by antigens an APC finds within the context of a troubled cell, whereas those merely escorted by soluble proteins typically produce an antibody response from the immune system’s B cells. However, the HSP-peptide complex anomalously stimulates not only precisely targeted cytotoxic T cells, but also the immune system’s nonspecific natural killer (NK) cells. “The latter are particularly adept at whirling into action on short notice and attacking foreign invaders as well as tumors,” says Srivastava, who is now director of the University of Connecticut-based Center for Immunotherapy of Cancer and Infectious Diseases as well as Antigenics’ chief scientific officer.

(Therapeutic monoclonal antibodies such as Genentech’s Herceptin or IDEC’s Rituxan are acting as drugs, not as immunostimulants: they bind to important disease-associated cell surface receptors, altering their behavior. HSPs, on the other hand, directly sensitize the immune response to specific antigens.)

All this has just recently penetrated the textbooks. Top tier, peer reviewed journals are bulging with papers on HSPs’ immunological prowess. John Sogn, deputy director of Division of Cancer Biology at the National Cancer Institute of the National Institutes of Health, described Antigenics’ research as the “nicest, best integrated, basic science, and clinical story there is in vaccine research,” calling it “an incredible science story that beautifully integrates the developments in tumor immunology.”

Says the Mayo Clinic’s Richard Vile: “From our own experiments and data, it’s obvious HSPs are big time. The only thing is how they’re working. My personal feeling is that HSPs will turn out to be a major mediator of immune recognition of tumors and perhaps other diseases as well.”

Antigenics is working hard to prove researchers like Vile right.

A Fairly Simple Complex Concept

Antigenics is a company with more than 200 employees that, with its acquisitions of Aquila Biopharmaceuticals and Aronex Pharmaceuticals, has significantly enhanced the depth and reach of its product portfolio. The company is tackling cancer on the premise that HSPs’ indiscriminate peptide binding lets you direct the immune system against a tumor, without ever needing to identify a single tumor-relevant antigen.

“Here’s how it’s done,” according to Neal Gordon, Antigenics’ senior vice president of manufacturing operations. “A tumor, after surgical removal and biopsy, is packed in dry ice and shipped off via overnight express (à la FedEx or Airborne) to the company’s 30,000-square-foot processing facility in Woburn, Massachusetts. There, the tumor cells are broken open. Their HSPs (complexed with tumor-specific peptides as well as a plethora of irrelevant antigens) are extracted, put into vials, frozen and shipped back when the patient has recovered from surgery.”

“One great virtue of Antigenics’ personalized approach,” says Chairman and Chief Executive Officer Garo Armen, “is that you don’t have to know which is the relevant antigenic peptide — it will be in there somewhere along with all the irrelevant ones.” A corollary implication is that the methodology potentially could work for all cancer types.

“There’s no question that heat shock protein immunotherapy is one of the most promising areas right now,” says Daniel Von Hoff, professor of medicine, molecular biology, and pathology at The University of Arizona and Director of the Arizona Cancer Center. Von Hoff, a past president of the American Association of Cancer Research and a founder of Ilex Oncology, a biotechnology company, says, “Many, many people in the field — including a lot of experts here at the Arizona Cancer Center — believe that with their HSPs, Antigenics has a unique approach to every tumor.”

Typically, patients receive 25 to 50 micrograms of HSP-peptide complex in an intradermal injection. This is done once a week for four weeks, then every other week as long as the supply lasts. Each tumor’s size places a limit on how much HSP can be extracted from it, but there is usually enough for treatment. Antigenics is also working on ways of expressing a tumor’s antigenic repertoire in a laboratory cell line, thus providing amplified amounts of tumor-specific HSP-peptide complex when needed.

Personalized immunotherapy isn’t cheap. Armen projects the cost for a course of treatment at somewhere between $10,000 and $20,000, but that is no more than the cost of current biologicals such as Herceptin and interleukins. Moreover, the low incidence of side effects make HSPs a bargain from both price and quality of life standpoints. Armen, who estimates the size of the US cancer market at $30 billion annually, says the Woburn facility can now process 10,000 patients a year, yielding $150 million in annual revenues.

In the Clinic

A successful IPO in February 2000 left Antigenics holding more than $90 million in cash at year’s end, thereby allowing the company to self-finance several trials of its HSP-based immunotherapy for a broad range of indications, including kidney cancer, melanoma and pancreatic cancer.

Kidney cancer

Farthest along are trials for renal cell carcinoma, the most common type of kidney cancer. With more than 30,000 new cases a year in the United States, renal cell carcinoma accounts for about 85 percent of all kidney tumors. The Phase III renal cell carcinoma trial of Antigenics’ HSP immunotherapy is ongoing at 150 centers worldwide, with more than 360 patients currently enrolled. In the trial, patients are randomized to either just surgery (the standard treatment) or surgery plus HSP treatment. The company expects to see interim results in late-2003.

“The way you treat patients with kidney cancer is to take the tumor out,” says Von Hoff, who helped design Antigenics’ Phase III kidney cancer trial. “But in a large percentage of those patients, the tumors come back. Right now there are no other therapies for early stage cancer. You do the surgery, and just watch them afterwards. There’s nothing else we have to offer in this situation.”

“The disease typically spreads to a lung or lymph node,” notes Robert Amato, an oncologist in the urology department at Baylor College of Medicine. “But its metastases can appear anywhere. And it doesn’t respond well to chemotherapy,” he says. “Once a tumor metastasizes, life expectancy averages less than one year. Conventional therapy — interleukin 2, the only drug ever approved for advanced kidney cancer, or interferon, which is used off-label — helps only about 12 percent of the time at best and causes troublesome side effects.”

Antigenics uses its Woburn facility to turn the tumor against itself. “Feasibility is not an issue,” says Amato, who oversaw Antigenics’ 42-patient Phase I and 78-patient Phase II trials in renal cell carcinoma. “Fewer than 10 percent of the patients were unable to have vaccine prepared for them, and the reason, almost always, was that we couldn’t get the tumor out. There was no major adversity, and the treatment’s activity was equal to that of any other single agent that’s been tried.” Nearly half of the Phase I patients were still alive two years after receiving the HSP-based vaccine. With conventional treatment such as it is, the standard two-year survival rate is 15 percent.

Melanoma

According to the American Cancer Society, melanoma only accounts for about 4 percent of skin cancer cases, yet it causes about 79 percent of skin cancer deaths. It is estimated that in 2002, there will be almost 54,000 new cases of melanoma in the United States, and about 7,400 people will die of the disease. Current treatment options are limited to surgical removal, radiation, chemotherapy, immune therapy or a combination of these treatments — all of which are associated with significant adverse effects. Phase III research of Antigenics’ HSP vaccine in advanced melanoma began in 2002.

“Antigenics’ vaccines have also elicited positive outcomes with advanced melanoma,” says tumor immunologist Giorgio Parmiani, deputy scientific director of the Instituto Nazionale Tumori (National Cancer Institute) in Milan, who was lead investigator of a four-center Phase II melanoma study evaluating 28 patients with advanced cancer incurable with surgery. Five patients responded favorably to HSP treatment, including two in whom all evidence of melanoma disappeared for almost two years, and one who is still free of disease.

“Safety problems once again were minimal,” says Parmiani, who presented details of the trial during annual meetings of the American Society of Clinical Oncology (ASCO).

Pancreatic cancer

Pancreatic cancer is the fourth leading cause of cancer death in men and women in the United States. The American Cancer Society estimates that 30,300 Americans will be diagnosed with pancreatic cancer in 2002, and that 29,700 will die of the disease.

The great majority of patients have advanced disease by the time they’re diagnosed. Once a tumor has spread beyond its original site, the prognosis is bleak. Typical median survival is on the order of four or five months in untreated patients; the average advanced pancreatic cancer patient receiving chemotherapy survives only an additional month or two. At best, one in five patients survives the first year. In a large study carried out by the National Cancer Institute, none of the 126 participants — all with late-stage pancreatic cancer treated by conventional chemotherapy — lived longer than 19 months.

In 1997, Antigenics initiated a 15-patient, Phase I pancreatic cancer trial at Memorial Sloan-Kettering Cancer Center, but could get material for the vaccines for only five of the subjects, said Jonathan Lewis, the former Sloan-Kettering surgeon who ran the study. Pancreatic enzymes, activated by surgery, were degrading the HSPs and the trial was suspended.

However, Antigenics researchers were excited to learn that more than three years after the trial suspension, two of the five subjects of the original trial were still alive, given that the historical median survival of pancreatic surgery patients is 16 months. One patient treated with Oncophage is alive and disease-free 33 months post-surgery; the remaining four died at 8, 17, 30 and 36 months after surgery. Antigenics’ scientists have now found a way to prevent enzymatic destruction of HSPs. Lewis, who left Sloan-Kettering to become Antigenics’ chief medical officer, said that the company has resumed the trial as a Phase II study after recruiting five new subjects.

Other cancers

Clinical results in other indications have also been encouraging. In a Phase II trial, results of which were presented at ASCO in 2001 and 2002, 29 patients with advanced colon cancer that had spread to the liver received Oncophage after liver surgery. Although the study was not designed to evaluate clinical effectiveness, a small group of patients with favorable prognostic factors who received Oncophage were cancer-free longer than expected.

Antigenics is also evaluating its treatment in a Phase I trial in gastric cancer. Sixteen metastatic gastric cancer patients, with life expectancies of approximately four to seven months, were treated in the study. After an initial evaluation, approximately half of the patients appear to be benefiting.

Research with the company’s HSP treatment also includes studies in non-Hodgkin’s lymphoma and soft tissue sarcoma. “Although none of these trials have used controls, survival rates are up significantly compared with historical rates for patients in similar condition at baseline,” said Lewis. “Importantly,” he adds, “in clinical trials, the treatment appears to be well tolerated in the 300-plus patients who have received Antigenics’ HSP immunotherapy.

“Antigenics’ clinical trials were preceded,” Lewis says, “by what may be the most extensive preclinical studies performed in the anticancer area. These studies involved greater numbers of animals within a wider assortment of species for more indications than have ever been reported with respect to any other cancer therapeutic,” he says, “and never with a negative result.” Lewis adds that the strong degree of homology between mouse and human HSPs, HSP receptors, and immune systems in general bode well for clinical trials.

Competition

Antigenics’ intellectual property portfolio, with its acquisitions of Aquila Biopharmaceuticals and Aronex Pharmaceuticals, contains more than 80 issued and 150 pending patents, including 36 carefully-focused, issued US patents, giving the company a lock on all immunotherapeutic use of noncovalently bound complexes of mammalian HSPs and antigens.

Antigenics’ clinical trials currently use the human HSPs called gp96 and HSP70, the moieties for which a specific receptor on dendritic cells has already been identified definitively. But its patents cover all the other members of the mammalian HSP family as well. Other cancer-vaccine contenders fall broadly into two groups. The first uses wide-band antigenic approaches, as Antigenics does. But instead of carefully extracting an active principle such as HSPs, competing variations on this approach by and large use whole cells (for instance, dendritic cells fused with tumor cells, altered to produce stimulatory cytokines, or infected with virus carrying genes for antigens). Intact cells require very careful handling. Their sterility is difficult to ensure, which poses a regulatory concern.

The second group consists of those inoculating with specific antigens believed to be widely shared among tumors of a particular type. A problem with single antigens is that they tend to differ from one tumor type to the next and often researchers are learning to their chagrin, that they also tend to differ within a given 'type.' Therefore, each indication requires the identification of an effective new antigen. “Moreover,” observes Parmiani, “tumor cells are constantly dividing and, therefore, constantly mutating. Antigens may be eliminated as the tumor grows, so an initially effective vaccine won’t work.”

“Attempts to find truly tumor-specific single antigens common to large numbers of patients have generally come to naught,” says Srivastava. Those so occupied have usually had to content themselves with antigens that were not unique to, but merely overrepresented on, tumor cell surfaces. These antigens are more likely to be tissue- rather than strictly tumor-specific. Researchers thus have had to steer a middle course between the Scylla of side effects inherent in attacking innocent cells displaying diminutive quantities of these antigens and the Charybdis of immunotolerance.

Indeed, Srivastava believes that each tumor’s in vivo immunogenicity may be inherently individualized. This immunogenicity, he suspects, is attributable to the idiosyncratic peptide products of random mutations that inevitably arise in undisciplined tumor cell populations. Of the myriad of other antigenic components present on a cell’s surface, even those drastically overrepresented in cancer cells are likely to have been rendered tolerogenic in early development. They may be found at lower levels on healthy cells as well, raising the possibility of side effects even if an approach succeeds in overcoming that immunotolerance.

There exists a class of exceptions to the rule that no single antigen a) is found on all tumors of a given type, b) never appears on healthy cells, and c) is nontolerogenic: antigens produced by pathogens’ oncogenes. Because they’re of foreign origin, they are capable of producing a strong immune response. One of Antigenics’ competitors has adopted a vaccination technique that employs a bacterial HSP covalently linked to a single antigen from a viral oncogene implicated in cervical cancer. Early clinical trials indicate some measure of success, although some researchers, such as the Mayo Clinic’s Vile, wonder whether serial injections of bacterial HSPs — the amino acid sequences of which bear similarities to those of mammals, but nonetheless differ significantly — might themselves trigger the formation of neutralizing antibodies at levels that could interfere with the vaccine.

From Gene to Vaccine?

Not only viral oncogene products, but also viral antigens in general, have the potential to trigger an immune response when properly presented to APCs. This means HSPs’ ability to deliver antigens to APCs is not limited to tumor antigens, but applies to a wide assortment of infectious diseases. Intracellular pathogens in Antigenics’ sights include herpes simplex virus 2 (HSV-2), HIV and tuberculosis, all of which successfully evade immune detection once they secure an intracellular beachhead. Immunologists are increasingly convinced that immunotherapeutic jump-starting of the cell-mediated response is key to combating such chronic infections.

The US Food and Drug Administration has approved clinical trials of an HSP complexed with a single, defined HSV-2 antigen (or a select few of them), and a Phase I trial is underway at the University of Washington in Seattle. Some 45 million Americans — a full 22 percent of the adult population — are infected with HSV-2, says Larry Corey, head of the University of Washington’s virology division and director of the school’s herpes clinic, where the trial is taking place. In this pilot study, investigators will be looking to see if vaccination enhances the cellular immune response that, Corey says “may be the most important element in controlling the disease.”

“Most pathogens exhibit much less variation than tumors and so don’t need to be tailored to individuals,” notes Armen. That suggests Antigenics could realize dramatic cost reductions by brewing up large batches of virus — for example, in an immortalized laboratory cell line or in intact organs — and harvesting the infected cells’ HSPs along with their assemblages of bound pathogenic antigens.

Armen sketches an alternative scenario: “In the wake of the Human Genome Project there lies a reserve army of surplus DNA sequencers. Some of that apparatus, left idle by the winding down of the program, could be easily diverted to the full tilt sequencing of numerous pathogen genomes. These could then be ‘filtered’ through computer algorithms predicting various amino acid sequences’ ability to bind both HSPs and common human MHC alleles. A manageable number of such peptides,” Armen suggests, “could be synthesized and bound in vitro to HSPs, creating a multivalent vaccine that was certifiably pure and quantitatively standardized. This would alleviate potential regulatory worries, if there were any, regarding the safety of extracts from lab cell lines. Such a vaccine’s multivalent character would nevertheless provide insurance against a pathogen’s mutating into an escape variant,” says Armen, “and should one arise, you’d just have to isolate the HSP-peptide complexes, not characterize the antigens.

“To the extent that Antigenics is successful in teaming HSPs with single antigens (or small numbers of them) against invading pathogens to boost infected individuals’ immunity,” Armen says, “continuing advances in genomic sequencing that yield a better understanding of MHC variations among humans will increase that success rate by allowing for fine-tuning of antigen choices.”

New Frontiers

Two separate developments have turned Antigenics scientists’ attention toward yet another disease category: autoimmunity. First is the observation that big doses of HSP-peptide complexes — five to 10 times the therapeutic amount — downregulate the immune response to those peptides. Srivastava’s University of Connecticut lab has had good results with this treatment in animal models of diabetes and multiple sclerosis. The second noteworthy development is the identification and ongoing characterization of CD91, the HSP receptor on antigen-presenting cells. Antigenics has begun a screening program to systematically search for small molecules that can modulate the receptor’s responsiveness to HSP-induced stimulation.

When Antigenics bought Aquila Biopharmaceuticals in 2000, it took ownership of seven corporate partnerships, a commercial feline leukemia vaccine and a general-purpose immunostimulatory compound, or adjuvant, that is far along in development. The adjuvant, QS-21, has been tested in more than 3,500 patients in over 50 clinical trials, and has shown to be more effective than alum (the only currently approved adjuvant) not only in inciting a cellular immune reaction but also in arousing the humoral branch of the immune system.

With the purchase of Aquila Biopharmaceuticals, Antigenics also inherited a highly significant additional piece of intellectual real estate to which Aquila holds title: an APC surface-based receptor called CD1. This receptor — actually a family of proteins related to MHC molecules — recognizes antigens that have lipid (fat-like), as opposed to peptide, components. CD1 then reacts by turning up the cellular response (on the part of both classic T cells and a specialized, immunoregulatory class of NK cell) to those antigens. Research strongly suggests that there’s a place for CD1 modulation — and for a major thrust in discovering whole new classes of suitable lipid-containing antigens — in fighting cancer, infectious disease and autoimmunity.

Antigenics’ acquisition of Aronex Pharmaceuticals in 2001 brought into the company two advanced, liposomal chemotherapeutics. The liposomal delivery system entraps drugs within shells of fat molecules, which could potentially increase distribution and duration of a medication within the body. The first liposomal product acquired in the merger is Aroplatin, a third-generation platinum agent that will enter Phase II research in 2002 in pancreatic and colorectal cancers. The second, ATRA-IV, is an intravenous formulation of an existing oral chemotherapy. Antigenics plans to test its liposomal formulation of ATRA, which is a form of vitamin A, in hematological malignancies in 2002.

While conventional drug discovery companies try to make heads or tails out of the vast torrent of new genomic information that has begun to rain down upon them, Antigenics finds itself in the enviable position of being able to skip lightly around functional genomics, a discipline of enormous potential that, however, promises to be expensive, time-consuming and, at least for a while, prone to missteps. In the short run, Antigenics can bypass the characterization of individual, relevant antigens and proceed full speed ahead with its wholesale attack on tumors. And in the long run, the company can take on not only cancer but also infectious disease and autoimmunity, using a variety of disparate but complementary approaches that will increasingly draw on the burgeoning of detailed information about individual antigens that genomic research will bring about. The steady advance of its HSP-based vaccine through clinical trials in a broad range of indications, along with the synergies wrought by its acquisitions of Aquila Biopharmaceuticals and Aronex Pharmaceuticals, suggests that Antigenics will be launching a sustained volley of products in high stakes arenas in the years ahead.

Bruce Goldman is a freelance scientific writer who lives in San Francisco.