FEATURE

The roller coaster

Bruce Goldman & Laura DeFrancesco

The field is littered with promising products that failed to show clinical efficacy. Could it finally be on the verge of a first US approval?

f any field epitomizes the boom and bust from the , which has antigens (Fig. 1 and Table 4). Many such tri- Icycles of biotech, it would be cancer . been burned repeatedly (Table 2). And what als have ended in failure, which we now know Over the years, numerous tumor immunother- lessons from the ever-growing list of failures— is because these antigens muster only weak apies have gone through rounds of early-stage and some possible successes—will inform immune responses because they are normal successes, only to fail in phase 3 clinical trials. future practitioners in the field? human proteins merely overexpressed on Experts point to many reasons for the failures, tumor cells (to which the patient would be from “jumping the gun” before enough was Beginnings tolerant) or they too closely resemble such known about the biology or the therapies to Cancer vaccinology is predicated on the proteins or they elicit only a weak response letting business considerations—going for low notion of awakening the to from the patient’s compromised immune cost and short time lines—trump science; what the presence of cancer by presenting it with system. It is now known that multiple co- Peter Bross, chief of clinical evaluations at the antigens associated with tumor cells. Once stimulatory signals are needed to generate a US Food and Drug Administration’s (FDA’s) the immune system is roused, the concept is robust T-cell response against a tumor-asso- Center for Biologicals Evaluation and Research that it would be capable not only of mount- ciated antigen; if these signals are not sup- calls companies simply not doing their home- ing a sustained bodywide search for simi- plied, T-cell anergy and peripheral tolerance work. Put these problems together with poorly larly suspicious cells, but also of retaining follows. Such tepid immune responses are designed clinical trials of heterogeneous cancer a memory of the abnormal antigens, per- not nearly what would be needed to eradicate patient populations with late-stage disease, add mitting a renewed, rapid assault should the advanced , which early on accounted a lack of familiarity of the regulatory authori- tumor recur. for most patients treated in clinical trials. ties in assessing tumor vaccine products, mix The notion that the immune system could Contemporary trials using tumor-associated in manufacturing scale-up headaches and the be enlisted to launch an attack on an exist- and more promising tumor-specific anti- resulting recipe is all but toxic to investors. As ing tumor has been around at least since the gens now use various immune stimulatory Bruce Booth of Atlas Ventures (Waltham, MA, late 1800s, when the New York City–based molecules, such as granulocyte macrophage USA) puts it, realizing the potential of cancer physician William Coley noticed that metas- colony stimulating factor (GM-CSF), and vaccines is “full of complexity.” tases at several sites regressed in a sarcoma generalized adjuvants, such as keyhole limpet But some researchers and analysts are patient after she developed a bacterial inci- hemocyanin (KLH), to boost the response. keeping the faith, hoping that a more com- sion-wound infection. Coley’s attempts to Many approaches have explicitly tried to prehensive understanding of tumor immu- exploit this discovery were handicapped by engage cell-mediated immunity either using nology will lead the way to more fruitful the then-crude state of knowledge. But to this isolated antigen-presenting cells (APCs) or approaches (Table 1). Several promising day, remnants of this approach can be seen attempting to stimulate them in situ (Fig. 2). phase 3 programs are nearing completion, in the use of general immune stimulants, like Techniques were developed for extracting so 2009 may well be the year of the cancer attenuated bacteria (e.g., mycobacterial com- dendritic cells, a major APC, loading them up vaccine. “There have been other technologies ponents in Bacille Calmette Guerin, BCG) with tumor antigens in various ways and rein- that failed in their first iteration…. As long as and interleukins, in treating bladder cancer troducing them into patients. Early attempts modifications are made and something new and , respectively, as well as their here failed, and in some cases, actually led to comes out of it, I think you’ll generate inter- inclusion in combination therapies in liter- poorer outcomes than if the individual had est,” says Reni Benjamin, senior biotech ana- ally hundreds of clinical trials. been untreated, as immature dendritic cells, it lyst at Rodman and Renshaw (New York). The discovery and identification of tumor- was later learned, were as likely to suppress the In the meantime, the question is whether associated antigens, which now number in immune system as to stimulate it. Methods for there is enough money to support the the hundreds (see Table 3 for some exam- characterizing the right types of dendritic cell approach in the coffers of biotechs or coming ples), stimulated a second approach to can- and other APCs are now being worked out, cer vaccines, an approach still highly visible and it’s become clearer how to activate these Bruce Goldman is a freelance writer living in among the therapies being tested today. cells through cytokines, such as GM-CSF, San Francisco, and Laura DeFrancesco is Senior Roughly half of ongoing clinical trials enlist to optimize antigen presentation (one such Editor, Nature Biotechnology. a tumor-associated antigen or collection of immunotherapeutic candidate in late-stage

nature biotechnology volume 27 number 2 FEBRUARY 2009 129 feature

Table 1 Selected early stage cancer vaccine programs Company (location) Product Composition Indication Trial phase Antigen Express Her-2/neu breast cancer Her-2/neu epitope peptide con- Breast cancer Phase 2 (Worcester, MA, USA; a subsid- vaccine jugated at N terminus to the C iary of Generex Biotechnology, terminus of the key moiety of the Toronto) MHC class II–associated invari- ant chain (Ii protein) containing a four–amino-acid (LRMK) modi- fication Apthera NeuVax Immunopeptide (E25) from Early-stage breast cancer Phase 1/ 2 (Scottsdale, AZ, USA) Her-2/neu administered together with GM-CSF Argos Therapeutics AGS-003 Autologous dendritic cells Renal cancer Phase 2 (Durham, NC, USA) loaded with total RNA from resected tumors Immunocellular Therapeutics ICT-107 Autologous dendritic cells Brain cancer Phase 1 (Los Angeles, CA, USA) treated with tumor-specific peptides from 6 antigens expressed on glioblastomas Immunotope IMT-1012 Peptide vaccine containing Advanced ovarian and breast Phase 1 (Doylestown, PA, USA) 12 tumor-associated peptides cancer discovered through proteom- ics, including A-kinase anchor protein 9, midasin (MIDAS- containing protein RAD50), talin 1, vinculin vimentin and cen- trosome-associated protein 350 Pevion Biotech Pevi-Pro Influenza virosomes expressing Breast cancer Phase 1 (Bern, Switzerland) three Her2/neu epitopes Vaxon Biotech Vx-001 A peptide vaccine comprising NSCLC Phase 1 (Paris) the cryptic peptide human telomerase reverse tran- scriptase (TERT572) and its HLA-A*0201-restricted modified variant (TERT572Y)

Table 2 Selected deals in the cancer vaccine sector Date Company (location) Partner (location) Product Deal terms Status 12/08 Oncothyreon Merck KgaA Stimuvax for NSCLC $13 million for manufac- Phase 3 turing rights 11/08 Argos Therapeutics Private investment RNA-loaded autologous $35.2 million for series C AGS-033 for renal cell dendritic cells (and other funding carcinoma in phase 1/2 ) 4/08 Cell Genesys Takeda GVAX for prostate cancer $50 million up front, plus Collaboration terminated (Osaka) milestones worth up to after GVAX trial stopped $270 million (12/08) 3/07 Oxford Biomedica aventis TroVax (allogeneic modi- $690 million in royalties Phase 2 vaccinations (Bridgewater, NJ, USA) fied vaccinia strain Ankara and milestones stopped due to excess expressing 5T4 (OBA1) deaths (7/08), analysis antigen) for renal cancer continues of vaccinated patients 12/04 CancerVax Merck/Serono Canvaxin for melanoma $25 million cash up front, Partnership terminated $12 million equity pur- (11/05); CancerVax chase; equally share devel- merged with Micromet opment costs, up to $253 in 2006 million in milestones 4/03 Accentia BiovaxID (autologous $20 million; Accentia Phase 3 idiotypic determinant from owns 81% of Biovest B-cell lymphoma conju- gated to KLH and com- bined with GM-CSF) for non-Hodgkin’s lymphoma 7/01 IDM Pharma sanofi aventis Uvidem (autologous den- $33 million Partnership terminated (Irvine, CA, USA) dritic cell vaccine loaded (1/08) ex vivo with tumor antigens derived from resected tumor) for melanoma

130 volume 27 number 2 FEBRUARY 2009 nature biotechnology feature clinical trials, Dendreon’s Provenge (Seattle; immunoprotective, prophylatic vaccine like sipuleucel-T) for prostate cancer, may prove Table 3 Examples of tumor-specific smallpox or polio where a viral antigen is to be among the first therapeutic cancer vac- antigens presented to the immune system. In those cines to receive FDA approval; Box 1). Tumor-specific, shared antigens cases where cancers overexpress a particular Just in the past five years, information has Cancer only endogenous surface antigen (e.g., Her-2 in surfaced, pointing to a whole new problem some breast cancers or CD-20 in some lym- with cancer —active immu- MAGE-3 phoma cells), mAbs directed against those nosuppression in the tumor microenviron- NY-ESO-1 surface markers (’s Herceptin ment. Tumors have been long suspected to TRAG-3 (trastuzumab) and Genentech’s and - evade immune detection by, for example, Expressed in some normal tissues Idec’s Rituxan (rituximab), respectively) Darwinian evolution of cells whose defining provide passive immunity, which can keep a WT-1 surface antigens are suppressed or creating tumor in check for a while. There are many positive pressure gradients that make it harder PRAME such mAbs for various cancers under devel- for circulating immune cells to penetrate SURVIVIN-2B opment. As currently applied, these mAbs them (Fig. 3). But now, it has emerged that Overexpressed in cancer are not preventive but rather therapeutic, in addition to evasion, tumors actually can though Herceptin has been approved for ever Her-2 induce local immunosuppression through earlier stages in breast cancer, where it might, the stimulation of regulatory T cells or the MUC-1 at least in theory, protect against recurrences recruitment of myeloid-derived suppressor Survivin by preventing metastases from taking hold. (MDS) cells. The former, primarily through Mutated, unique Active immunotherapies, on the other their production of transforming growth fac- p53 hand, are designed to incite the individual’s tor (TGF)-β, inhibit CD8+ cytotoxic T cells own immune system to mount a response to α-actinin-4 (CTLs), T helper 1 (Th1) cells and natural an antigen or group of antigens exclusive to or killer (NK) cells, which are the main media- Malic enzymes predominantly associated with the patient’s tors of immune surveillance against tumors. Source: GSK tumor. They can take the form of peptide/ MDS cells, a mixed population of relatively protein vaccines or cellular vaccines. immature myeloid cells, also suppress cellular The former type of vaccine generally falls immune responses primarily by producing into two categories. The first is based on arginase 1 and nitric oxide synthase 2A. suppression. Indeed, several dozen clinical tri- shared peptide or protein antigens that occur One means of potentiating the power of als, according to the US National Institutes of commonly in a particular cancer or group cancer vaccines and unleashing the immune Health (http://www.clinicaltrials.gov), are cur- of cancers (epidermal growth factor recep- system, according to leading academics, would rently underway using mAbs against CTLA-4 in tor (EGFR) vIII, for example, which is found be to counteract tumor-mediated immune combination with chemotherapy or vaccines. in 30–40% of glioblastomas, or MAGE-3, suppression. This could be accomplished by which is expressed on many lung tumors). targeting the regulators of the regulators, so Immunotherapy’s many faces The proteins can be injected directly or to speak. For example, several molecules have means different expressed on attenuated virus particles, or been identified (e.g., CTL antigen 4, CTLA-4) things to different people. In the case of nonproliferative bacterial or yeast cells (Box that engage with regulatory T cells. Animal cancers that are known to express viral anti- 2). An alternative approach is to isolate studies have shown that blocking such inter- gens (e.g., cervical cancer and some melano- antigens from an individual patient and actions, either with monoclonal antibodies mas that express human papilloma virus), present these back to the person in a form (mAbs) or gene knockouts abrogates immune immunotherapy takes the form of a classic designed to elicit immune surveillance, such

60 35

50 30 25 40 20 30 15 20 10 Number of products Number of products 10 5

0 0 Phase 1 Phase 2 Phase 3 s ticle ccines ydrates mRNA va Cell-based e par Antibodies DNA Carboh Protein antigens us-lik Synthetic proteins

us or vir Vir

Figure 1 Cancer vaccine types. (Source: Tufts Center for the Study of Drug Development)

nature biotechnology volume 27 number 2 FEBRUARY 2009 131 feature as vaccines designed to stimulate responses type of vaccine would be a product based 3 trial of cellular vaccine Canvaxin because against antibody idiotypes found on lym- on isolation of APCs from a patient that is of manufacturing concerns. What’s more, the phomas or the use of heat shock proteins to engineered to express some soluble factor (or longer clinical history and widespread use of present unique tumor peptides (Box 3). factors) that generates an immune response peptide/protein vaccines means that regula- Cellular cancer vaccines can also be to a common antigen (e.g., prostate-specific tors are more familiar with their oversight divided into two broad groups: allogeneic antigen in the case of prostate cancer, or p53/ and less likely to raise issues unanticipated or autologous. The former, so-called ‘off- telomerase more generally (Box 1)). by product sponsors. the-shelf’ vaccines, are usually collections of Compared with cellular vaccines, peptide tumor cell lines, administered as aggregates vaccines have the advantage of being similar to The perilous path to present several potential tumor antigens existing vaccine approaches used for decades Cancer vaccines represent a relatively small to the patient’s immune system. Autologous in immunization programs against infectious portion of the oncology drugs in commercial whole cells, on the other hand, are isolated agents. Such vaccines are less tricky to manu- development. The Tufts Center for the Study of from, and returned to, individuals after some facture on a large scale than cellular vaccines. Drug Development (Boston) reports that only ex vivo manipulation to activate or induce In 2002, for example, the FDA placed a hold one-fifth of oncology biologic therapeutics maturation of APCs. An example of this on CancerVax’s (Carlsbad, CA, USA) phase in company pipelines are vaccines (Fig. 4).

Table 4 Selected cancer vaccines in late clinical trials Company (location) Product Description Indication Trial phase Whole-cell-based autologous cells (personalized) M-Vax Autologous cell vaccine in which Metastatic melanoma with at Phase 3 (Philadelphia) patient tumor cells are treated least one tumor to create vaccine with the hapten dinitrophenyl Dendreon Provenge Autologous dendritic cells Asymptomatic, metastatic hor- Phase 3 exposed ex vivo to fusion protein mone-refractory prostate cancer combining prostate alkaline phosphatase and GM-CSF Geron GRNVAC1 Autologous dendritic cells trans- AML in remission Phase 2 (Menlo Park, CA, USA) fected with mRNA for human telomerase and a portion of lysosome- associated membrane protein (enhances antigen pre- sentation) IDM Pharma Bexidem Autologous interferon-γ-activated Superficial bladder cancer Phase 2/3 macrophages (monocyte-derived activated NK cells). Uvidem Autologous dendritic cell Melanoma with M1a or M1b Phase 2 vaccine loaded ex vivo with stage disease and/or in-transit tumor antigens derived from lesions; stage III and IV resected tumor melanoma Collidem Colorectal cancer Phase 1/2 Introgen Therapeutics INGN 225 Dendritic cells treated with an Advanced metastatic SCLC Phase 2 (Austin, TX, USA) adenovector carrying the human Breast p53 gene MolMed M3TK T cells bioengineered to express Metastatic melanoma Phase 2 (Milan) MAGE 3 tumor antigen (enrollment halted) Northwest Biotherapeutics DC-Vax Prostate Dendritic cells loaded with Hormone-dependent, nonmeta- Phase 3 (Bethesda, MD, USA) recombinant prostate-specific static prostate cancer membrane antigen (PSMA) DC-Vax Brain Dendritic cells loaded with Newly diagnosed glioblastoma Phase 2 tumor extract multiforma requiring surgery, radiation and chemotherapy Prima Biomed CVac Dendritic cells primed with a Late-stage ovarian cancer Phase 2 (Sydney, Australia) mucin-1 and a mannan-fusion protein adjuvant Whole-cell-based allogeneic tumor cells (off-the-shelf) Cell Genesys GVAX pancreatic Two allogeneic cultured cancer Metastatic pancreatic cancer Phase 2 lines, irradiated and bioengi- neered to secrete GM-CSF. GVAX leukemia One allogeneic leukemia cell line Newly diagnosed AML, chronic Phase 2 irradiated and bioengineered to CML and myelodysplastic secrete GM-CSF syndrome NovaRx Lucanix Four non-small cell Advanced NSCLC Phase 3 (San Diego) cell lines carrying antisense oli- gonucleotides against transform- ing growth factor β-2

132 volume 27 number 2 FEBRUARY 2009 nature biotechnology feature

Table 4 Selected cancer vaccines in late clinical trials (continued) Company (location) Product Description Indication Trial phase Onyvax Onyvax-P Three human cell lines repre- Hormone-resistant prostate Phase 2 (London) senting different stages of pros- cancer tate cancer Unique-antigen-based (personalized): purified peptide or protein Antigenics HSPPC-96 Oncophage Heat shock protein vaccine puri- Recurrent glioma Phase 2 (investigator- fied from autologous tumor cells initiated trial) Resected renal-cell carcinoma Phase 3 (completed) (RCC) Biovest International BiovaxID Tumor-specific idiotype conju- Mantle cell lymphoma Phase 2 gated to keyhole limpet hemo- Indolent follicular B-cell non- Phase 3 cyanin, plus GM-CSF Hodgkin’s lymphoma Shared antigen (off-the-shelf): purified protein or peptide Apthera NeuVax Immunogenic peptide derived Early-stage Her-2-positive breast Phase 2/3 (Scottsdale, AZ, USA) from the Her-2/neu protein plus cancer GM-CSF CellDex CDX-110 A 14-amino-acid segment of a Glioblastoma multiforme Phase 2/3 mutated EGFR Cytos Biotechnology CYT004-MelQbG10 Modified fragment of the Advanced-stage melanoma Phase 2 (Schlieren, Switzerland) Melan-A/MART-1 protein coupled to the carrier QbG10 Generex Biotechnology Ii-Key/HER2/neu cancer vaccine Peptide vaccine containing Node-negative breast cancer Phase 2 Ii-Key modified Her-2/neu pro- tein fragment GlaxoSmithKline Biologicals MAGE-A3 antigen-specific can- Liposomally packaged cancer Metastatic MAGE-A3-positive Phase 3 (Brussels, Belgium) cer immunotherapeutic vaccine against MAGE-3 antigen melanoma Phase 3 NSCLC following surgery IDM Pharma IDM-2101 Nine CTL epitopes from four NSCLC Phase 2 tumor-associated antigens, including two proprietary native epitopes and seven modified epitopes and one universal epitope (a source of T-cell help) Immatics Biotechnologies IMA901 Peptide vaccine comprising Renal cancer Phase 2 (Tuebingen, Germany) IMA910 multiple fully synthetic tumor- Colorectal cancer Phase 1/2 associated peptides Norwood Immunology Melanoma cancer vaccine Melanoma-specific peptides Melanoma Phase 2 (Chelsea Heights, Australia) gp100 and MAGE-3 Oncothyreon Stimuvax Liposomal vaccine containing a Stage lll NSCLC Phase 3 synthetic 25–amino-acid-peptide sequence from MUC-1 Pharmexa GV1001 Recombinant protein vaccine tar- Pancreatic Phase 3 (Hoersholm, Denmark) geting human telomerase reverse Liver Phase 2 transcriptase, plus GM-CSF Lung Phase 2 AML, amyotrophic lateral sclerosis; CML, chronic myelogenous leukemia.

Although modern cancer vaccine develop- • PANVAC (Therion Biologics, Cambridge, off-the-shelf vaccine, consisting of a synthetic ment dates back to the 1980s, none has been MA, USA), an off-the-shelf vaccine con- mimic (STn-crotyl) of the tumor-associated, approved in the United States (though there sisting of attenuated poxvirus carrying O-linked epitope of MUC-1 (STn-serine), are five products on the market elsewhere; genes encoding two tumor-associated tethered to an immunostimulatory protein Table 5). Thus, the rate of approval of cancer antigens (carcinoembryonic antigen and (KLH) and delivered along with an adjuvant vaccines lags far behind other biologics—as mucin 1, MUC-1) and three immunos- from Seattle-based Corixa (Detox-B, an oil of 2006, seven of twelve vaccines in phase timulatory molecules (intracellular adhe- droplet emulsion containing monophos- 3 clinical trials had entered clinical study a sion molecule 1, B7.1 and lymphocyte phoryl lipid A and cell wall skeleton from decade earlier. function–associated molecule 3) for use Mycobacterium phlei) for use in metastatic To date, an estimated 7,000 people have in advanced pancreatic cancer, failed to breast cancer, showed no improvement in participated in late-stage clinical trials of meet clinical endpoints after promising either time to progression or overall survival. active cancer immunotherapies. These have early trials, leading the company to close The company hasn’t completely abandoned largely been an exercise in frustration, as its doors and file for bankruptcy protec- the target; in partnership with Merck KGaA candidates—including a few that looked tion in December 2006. (Darmstadt, Germany), it has developed a quite good in early trials—have fallen by the “more sophisticated” approach for elicit- wayside in pivotal phase 3 trials. Some recent • Theratope (Biomira, Edmunton, AB, ing a T-cell response, according to Marita losers that have gone quietly into the night: Canada; now Oncothyreon, Seattle), an Hobman, director of intellectual property

nature biotechnology volume 27 number 2 FEBRUARY 2009 133 feature

merged with Micromet, which is develop- Dendritic cell matures and is ing passive immunotherapies using mAbs infused back into against various tumor antigens. Tumor patient antigen is • GVAX (Cell Genesys, S. San Francisco, CA, linked to a Complex binds Complex is cytokine to dendritic cell take in by Tumor antigens USA), an off-the-shelf, whole-cell vac- precursor dendritic cell cine, consisting of infusions of cells from precusor existing prostate cancer lines engineered

T cell to express GM-CSF for use in hormone- Cancer cell refractory prostate cancer, yielded excess deaths in treated patients versus controls, Dendritic cell leading to abandonment of the trial. displays tumor antigen and Although there is a clear preponderance of activates T cells off-the-shelf vaccines in this group of failures, T cells attack cancer cell the fate of individualized vaccines has not nec- essarily been much better. Two companies with Figure 2 Dendritic cells that attack cancer. (Source: National Cancer Institute) vaccines targeting antibody idiotypes associ- ated with tumors—Favrille (San Diego) and Genitope (Fremont, CA, USA)—both shut management and business development at plus an adjuvant (BCG) for use in stage down their trials when their products failed Oncothyreon. III melanoma, yielded worse outcomes in to reach statistical significance, essentially end- treated patients than in controls, unlike ing their programs in late 2008. • Canvaxin (CancerVax, now MicroMet, earlier trials in which patients had been Munich), an off-the-shelf mix of three irra- more carefully selected for human leukocyte Getting it right diated melanoma cell lines bearing over a antigen (HLA) alleles correlating with better A cancer vaccine has to jump through sev- dozen defined tumor-associated antigens, outcomes. After Canvaxin failed, CancerVax eral hoops, says Johns Hopkins University

Box 1 A whole-cell vaccine nears approval?

Dendreon is developing a whole-cell-based candidate, Provenge, be a reasonable surrogate for survival. But Dendreon was to find for metastatic, hormone-resistant, prostate cancer (HRPC). The out otherwise. During the course of the trial, medical opinion vaccine is a patient-specific, vaccine produced by incubating leaders decided that overall survival is a better endpoint than an individual’s own blood, enriched for dendritic cells and other TTP, which carries a subjective component. However, the trial APCs with a recombinant fusion protein composed of prostatic continued with the previously agreed upon endpoint of TTP. acid phosphatase (PAP) and GM-CSF. This proved ironic. When the first trial was unblinded, TTP Although not tumor specific, PAP is highly tissue specific. had missed statistical significance (P = 0.05) by the barest Although expressed in the majority of prostate tumors, PAP of margins, (P = 0.052) whereas overall survival analysis is only minimally expressed in tissues other than the prostate demonstrated a 41% reduction in the risk of death, with a high gland, says Dendreon’s Frohlich, who is chief medical officer. level of statistical significance (P = 0.01). It is immunologically distinct from acid phosphatases found in But survival was not a prespecified primary endpoint. And other tissues. Because HRPC patients’ prostates have already whereas an outside advisory committee voted 13–4 for approval, been surgically removed or irradiated, autoimmunity doesn’t in April 2007, the FDA instead insisted on another trial to pose much of a practical problem. confirm the survival results. The question of whether tolerance to this antigen can be In the aftermath of the FDA’s failure to approve Provenge broken was addressed in a preclinical study performed by despite clear signs of efficacy, angry patient advocates peppered Dendreon investigators and published in 2001 (ref. 2), in the agency with letters of protest. But proponents of strict which their product induced autoimmune prostatitis in rats (a adherence to trial protocols liken the argument that a therapy clear sign of immune mobilization against the antigen). This ought to be approved on the basis of an unplanned analysis to convinced Dendreon that it could raise an immune response in moving the goal posts4. a clinical setting as well. A phase 1/2 trial published in 2000 Dendreon is soldiering on with a new 512-patient trial with the (ref. 3), demonstrating strong antigen-specific T- and B-cell primary endpoint of survival. Preliminary results, announced in responses to the approach, was consistent with this finding. October 2008, demonstrated a 20% reduction in the risk of death That year, Dendreon launched two trials of Provenge, each in the treatment arm, only slightly less than the 22% reduction, with about 120 asymptomatic patients. As Frohlich explains, it which the company believes is necessary to achieve statistical was then believed that asymptomatic patients would progress significance. Final results are due later this year, and if Provenge more slowly than symptomatic patients, buying time for the makes the grade, it may yet turn out to be the first whole-cell- initially subtle effects of immunotherapy to kick in before the based active cancer immunotherapy approved by the FDA. But disease reached a stage that was intractable to immunotherapy. many years, many tens of millions of dollars, and perhaps more In the interest of getting a fast readout, Dendreon had picked than a few lives might have been saved had the Dendreon’s phase as its primary endpoint time to progression (TTP)—assumed to 3 trial not been marred by an unfortunate choice of an endpoint.

134 volume 27 number 2 FEBRUARY 2009 nature biotechnology feature

oncologist Hyam Levitsky, co-inventor of Tumor-specific T cells GVAX and member of the board of the can- cer vaccine company Antigenics (New York). Tumor down-regulates “In an existing tumor, the body has already surface molecules been exposed to those antigens, so there may already have been an initial immune No recognition response. But very often, the immune sys- Microvesicles tem is defeated and rendered tolerant to the antigens that the vaccine is targeting. Cytokines, soluble A successful vaccine has to overcome this factors or microvesicles Tumor modulate systemic tolerance, and that’s not trivial.” Moreover, immune responses Levitsky says, the vaccine frequently has to work in what can be a hostile environment. Soluble “The tumor has essentially taken over and inhibitory factors altered the landscape, stealing various attri- CD4+/CD8+ butes of the normal immune system to turn T reg down immune response.” – Inhibitory – receptors/ligands The antigens to use in a vaccine to cir- – Immune cells in the CD4+ cumvent the challenge of breaking immune Dendritic cell tumor microenvironment tolerance without generating autoimmunity are functionally impaired CD8+ Th should be tumor specific. But such antigens are rarely found, says Jeffrey Weber, head CTL of the Comprehensive Melanoma Research Center at the H. Lee Moffitt Cancer Center (Tampa, FL, USA). “These are few and far between. You can discover any number of Figure 3 Tumor cell’s interactions with the immune system. (Reprinted from Whiteside, T.L. Immune mutated, tumor-specific antigens, but you suppression in cancer: effects on immune cells, mechanisms and future therapeutic intervention. seldom find any that turn up on more than Semin. Cancer Biol. 16, 3–15, 2006, with permission from Elsevier.) 5% of tumors of any given type.” And even when you find one, he says, that doesn’t mean it will be highly immunogenic. as no really tumor-specific melanoma anti- be observed more quickly in advanced-stage In practice, cancer antigens targeted by gens have yet been exploited, only tumor- patients than in early-stage or fully resected active immunotherapies have more often associated antigens. But those cell-based ones. But decades’ worth of clinical trials of been tumor associated: overexpressed on approaches, in which autologous proteins cancer vaccines conducted across multiple tumors, but nonetheless present at lower or extracts are used for priming, require tumor types not surprisingly suggest that frequencies in normal tissues. In trials of access to a sufficient tumor mass. This more immunotherapies are more likely to work vaccines based on these antigens, the neces- or less excludes melanoma or even breast best in patients with earlier-stage, less-aggres- sity of breaking tolerance—for example, by cancer, where the tissue tends to be fibrotic sive tumors1 or in individuals whose tumor pairing the selected antigen with a powerful and where tumors tend to be diagnosed burden has been reduced to the microscopic adjuvant—has clashed with the need to avoid increasingly early, while they are still rela- level by surgery or chemotherapy. an excessive immune assault on healthy tis- tively small. “It’s at this level of microscopic disease where sues where the antigen also resides. “You can Recognizing that the immune response I think cancer vaccines are most likely to suc- vaccinate the hell out of somebody against takes time to develop, some vaccine devel- ceed,” Levitsky says. “Well over 50% of the melanoma self-antigens that are overex- opers have turned to slow-growing prostate common cancers can be treated into a state of pressed on cancer, and you won’t induce cancer or kidney tumors, where the time to minimal residual disease. What we lose patients severe —or any immune response progression is longer. And then, of course, to is typically not the inability to get the disease to speak of,” says Weber. “But if you admin- greater prevalence of certain tumor types, into that minimal state, but rather the inability ister the same vaccine along with one dose such as lung, create a large patient pool with to completely eradicate the residual compo- of anti-CTLA-4 antibodies, you can induce which to populate clinical trials, whereas the nent.” All too often, a seemingly excised tumor life-threatening autoimmune colitis or skin dearth of decent treatments for these indi- returns. “From a public-health point of view,” he rash or hepatitis.” cations speaks most loudly to the need for says, “the impact of an effective immunothera- Another problem plaguing trials of cancer ramped-up clinical experimentation. py—delivered at the point of minimal residual immunotherapies has been the intractabil- Certainly, the tendency to use individu- disease—that could wipe out the last traces of ity of the cancers targeted. In theory, any als who are in advanced cancer stages has a tumor, would be truly staggering. Ironically, cancer should be amenable to immuno- made proof of clinical efficacy more difficult that’s probably the most difficult time to dem- therapy, but in practice, only a few cancers to achieve. Of course, individuals with late- onstrate efficacy in a .” have received most of the attention, at least stage disease, who have often been treated Standard measures of a cancer therapy’s effi- historically. Melanoma, which early on was with other therapeutic agents that have failed, cacy—tumor shrinkage or growth arrest—are found to have tumor-specific antigens, has tend to be more available. And sponsoring worthless for patients with minimal residual been targeted frequently using the protein or companies prefer this population because disease. How can you score tumor shrinkage if peptide approach—mostly without success, they expect that positive treatment effects will the patient no longer appears to have a tumor?

nature biotechnology volume 27 number 2 FEBRUARY 2009 135 feature

Box 2 Pharma perseveres with off-the-shelf vaccines

Off-the-shelf vaccines have the advantage that they can be Avant and are working on a large phase 2/3 trial, with produced in bulk, making them more attractive to big pharma preliminary results expected by mid-2009. With the aggressive than individualized vaccines that are tailored to each patient. One lethality that characterizes glioblastoma, those results shouldn’t such vaccine, CDX-110, developed by John Sampson at Duke be long in coming. “It’s not going take 20 years to find out if the University (Durham, NC, USA) and Amy Heimberger at MD Anderson vaccine worked,” says Heimberger. Hospital (Houston), has been in-licensed by Celldex Therapeutics Another vaccine in the pipeline that targets an antigen that is (Phillipsburg, NJ, USA, which merged with Avant Immunotherapics, both tumor specific and present in enough patients’ tumors to Needham, MA,USA, in late 2007) and has attracted the attention of allow an off-the-shelf, mass-produced vaccine is GlaxoSmithKline Pfizer (New York). (GSK) Biologicals (Brussels) MAGE-3 (melanoma antigenic epitope The vaccine targets EGFRvIII (a 14-amino-acid segment of a 3). Since October 2007, the vaccine-producing division of the mutated EGFR) that not only appears solely on glioblastoma cells, pharmaceutical giant has been recruiting non-small-cell lung cancer but also has never been expressed in any other kind of cell at any (NSCLC) patients for a large phase 3 clinical trial. The 400-center time in development. Its biological activity is clearly germane to the study, spanning 33 countries, will be the largest-ever in lung cancer tumor’s aggressiveness, so knocking it out should directly impair the and, for that matter, the largest-ever study of an active cancer tumor’s viability. It is located on cell surfaces, making it accessible immunotherapy for any indication. to attack. And, because it’s a mere peptide rather than a full-sized Vincent Brichard, senior vice-president for cancer protein, it’s simple to manufacture. immunotherapeutics at GSK, says that about 40% of all NSCLC EGFRvIII is found on 30–40% of glioblastomas, an aggressive tumors express MAGE-3, which is expressed only transiently during form of brain cancer, which even when surgically excised, irradiated fetal development. In adults, MAGE-3 expression is confined to the and exposed to chemotherapy, typically recurs within six months. testes, opaque to immune surveillance so that tolerance doesn’t The mutant receptor is characterized by a 267–amino acid deletion develop. within its extracellular domain, which changes the molecule’s This makes it possible to conceive of an off-the-shelf configuration, locking it into a perpetual signaling mode that drives immunotherapy targeting the widely shared antigen. The GSK relentless cell replication. Thus its 100% tumor specificity: no cell approach uses the entire 360 amino acid–long MAGE-3 protein to with elevated expression of this mutant receptor could possibly be maximize the number of epitopes. GSK’s vaccine bolsters the T-cell normal. In addition, the deletion creates a novel splice junction, immunogenicity of its recombinant MAGE-3 protein by packaging which CDX-110 spans. it in liposomes, which Brichard says enhances delivery to APCs The dearth of effective therapies for glioblastoma makes it and by administering the vaccine with an adjuvant mix that has possible to test the new vaccine as a front-line therapy in patients been optimized to produce a potent T-cell response to MAGE-3. who have just had their tumors thoroughly resected. In a phase 2 This immunostimulatory potion combines GSK’s own adjuvant, trial (ACTIVATE), 22 patients with EGFRvIII-positive glioblastomas monophosphoryl lipid A (MPA), with QS-21, a complex lipid mix were given standard treatment, followed by serial injections of the licensed from Antigenics, and CpG oligonucleotide (a Toll-like vaccine. Time to progression (TTP) more than doubled to more than receptor 9 agonist developed by Coley Pharmaceutical Group, a 14 months compared with 6.4 months for a set of EGFRvIII-positive Canadian biotech purchased in late 2007 by Pfizer). historical controls. Overall survival improved commensurately, QS-21 is also known to induce a strong antibody response. proving surprisingly enduring, with about two-thirds of the injected Antibodies could conceivably play a role against even an intracellular patients surviving for two years, more than one-third for at least three molecule such as MAGE-3 to the extent that lysed tumor cells years and a fifth still alive after four years. In a second phase 2 study release entire, undegraded protein molecules that could be targeted (ACT II), in which 23 subjects received the vaccine simultaneously by antibodies. The resulting antigen-antibody complexes would, in with chemotherapy, patients’ median TTP reached 16.6 months, and turn, be highly available for uptake by APCs. median survival time 33.1 months—a point at which, historically, all At the annual meeting of the American Society of Clinical EGFRvIII-positive patients would long since have died. Oncology in Chicago in June 2008, GSK presented the results of This jump in long-term survival is puzzling, as EGFRvIII-positive a randomized, 182-patient phase 2 trial of MAGE-3. Early-stage glioblastomas typically contain large numbers of EGFRvIII- NSCLC patients who had had their tumors completely resected and negative cells—a status that ought to shield them from the then received several injections of the MAGE-3 vaccine had roughly vaccine. One possible explanation says Sampson, is that EGFRvIII- one-third the recurrence rates of those given placebo injections, positive cells are stem cells for the tumor, another possibility is results mirroring those from several other MAGE-3 tests. that EGFRvIII-positive cells either make other tumor cells more GSK’s huge phase 3 trial—whose primary endpoint, like that proliferative or harder to kill. Chuck Baum, senior vice-president of the recent phase 2, is disease-free survival—departs from and head of the oncology development at Pfizer (New York), its phase 2 counterpart in two respects: first, about half of the which licensed the rights to the vaccine in April, 2008, suggests patients in the trial will receive chemotherapy before vaccination, that part of the apparently powerful effect CDX-110 has on even a regimen never tested in phase 2 (Brichard notes, though, that EGFRvIII-negative tissue may be due to a phenomenon called the trial’s large size leaves plenty of statistical power for an ‘epitope spreading’: as tumor cells lyse, they release their internal independent analysis of those eschewing chemotherapy). Second, contents into the surrounding medium, giving local APCs access there has been a change in the adjuvant mix’s composition—the to previously occult tumor antigens (e.g., mutant intracellular addition of CpG —since the phase 2 trial. Brichard says the new proteins). “You can get a broadening response with time, and mix proved more immunostimulatory compared with the earlier eventually end up with a more effective immune response than the formulation in another head-to-head trial, but it is into such small one you started with,” Baum says. cracks that surprises can flow.

136 volume 27 number 2 FEBRUARY 2009 nature biotechnology feature

An alternative is to monitor recurrences or, Whereas the failed Biomira vaccine, divergent timelines due to differential prognoses more accurately, deaths among treated versus Theratope, is an example of a highly purified, in different indications. untreated patients. But that can take a long time. well-defined antigen with a single epitope, “Any therapy that’s totally novel and first In renal-cell carcinoma, for example, the median CancerVax’s Canvaxin candidate was a whole- in class is a double-edged sword,” says Mark time to recurrence for patients who have had cell mixture with multiple antigens, some of Frohlich, senior vice president of clinical affairs their tumors fully resected and show no signs them undoubtedly not even identified, let alone and chief medical officer of Dendreon. “On of residual tumor is 6.8 years. characterized. The former approach runs the risk the one hand, there’s a lot of excitement from Further complicating cancer immunology of eliciting too narrow an immune response. The patients and regulators who want to approve trials is the fact that each approach tends to be latter may trigger unwanted cross-reactions, says something that’s new and different. On the other novel, creating trial-design and regulatory issues. Weber, and the difficulty of assessing its potency hand, if it’s a new product, those same regulators A designated antigen can be either tumor associ- in any given person poses regulatory issues. also need to make sure they’re doing everything ated or tumor specific, and tumor-specific anti- The roughly 30 different active cancer immu- to ensure public safety and establish a solid prec- gens can be shared by many patients or unique notherapies now in late-stage clinical trials edent.” Frohlich speaks from experience, hav- to each patient. Shared antigens offer the pros- (Table 3) also differ in their methods of manu- ing undergone an epic regulatory ordeal with pect of off-the-shelf vaccines, with attendant facture and the indications for which they’re Dendreon’s cell-based prostate-cancer vaccine economies of scale. But they also often lack being tested. Trial designs differ greatly, too. Provenge (Box 1). tumor specificity, thus incurring the drawbacks Among the variables: early- versus late-stage dis- For all these reasons, clinical trials of active of immune tolerance. ease trade-offs, different endpoints and widely cancer immunotherapies are high-stakes

Box 3 Personalized vaccines—a viable option?

Oncophage, a personalized vaccine developed by Antigenics, treatment began. Excluding these patients from the analysis consists of an extract containing heat-shock protein-peptide diluted the power of the study, which would have run much complexes prepared from an individual patient’s excised tumor. longer had these classification errors not been made. “One of This approach is based on work by company co-founder Pramod the reasons the trial did not meet its endpoints relates to the Srivastava, now director of the University of Connecticut’s fact that the data were evaluated prematurely,” says Christopher Center for Immmunotherapy of Cancer and Infectious Diseases Wood of MD Anderson, who was the principal investigator and (Farmington, CT, USA), showing that APCs have receptors for author of the Lancet paper. heat shock proteins. This provides a pathway whereby unique In March 2007, study investigators conducted a second tumor-specific antigens (the products of random mutations in analysis of more mature data, using an RCC classification rapidly dividing cancer cells) could become immunogenic. In system that hadn’t existed when the Oncophage trial had principle, this preparation can target any tumor type, but in begun. Vaccinated patients classified as “intermediate stage” practice, its application is limited to tumors of sufficient size to in this new system (a subset consisting of stage 1–3b, which obtain enough material. overlapped but was not equivalent to the older system’s “early In several early-stage trials involving individuals with stage” group) enjoyed a statistically robust (P = 0.026) relapse- advanced disease over several indications, there were striking free survival benefit, suffering recurrences at just over half the cases of complete tumor regression as well as instances of rate of untreated patients—as well as a trend toward statistically partial shrinkage or stable disease, although the overall results significant improvement in overall survival (P = 0.126). were unspectacular. Given the virtual absence of side effects, “What needs to be done, and hopefully will be done,” Wood Antigenics forged ahead, launching phase 3 trials in advanced says, “is another trial that focuses on that earlier-stage group. metastatic melanoma5 and renal-cell carcinoma (RCC)6. But the amount of money to do that would be enormous, The melanoma trial proved difficult, as excised tumors were because you’re narrowing down the population even further, so often too small to produce enough vaccine and subsequent it would be hard to accrue. And because they’re earlier-stage clinical development was halted. But subjects with early- disease, recurrences are less frequent, so it could take ten years stage disease who got ten or more injections saw a big survival before you reach statistical significance.” improvement over controls receiving currently approved The direct costs of the seven-year phase 3 trial exceeded treatments. $60 million, says Garo Armen, Antigenics’ CEO. “That, plus the In the RCC trial, initiated in 2000, more than 700 patients necessary maintenance of our manufacturing, quality control, whose tumors had been resected were randomized to either quality assurance and manufacturing-related research functions Oncophage or the current standard of care, which consists of throughout the trial, came to $250 million.” ‘watchful waiting’ as there are no approved, effective treatments Armen says about 500 trial patients will continue to be for such patients. As reported in The Lancet last July6, an followed up for relapse-free survival and overall survival for analysis triggered in November 2005 by the accumulation of a another three years, at a cost to the company of $1.5 million. specified number of inidivduals whose disease had progressed Meanwhile, in April, 2008, Oncophage was approved in found no statistically significant difference between treated Russia, where a large number of the phase 3 patients had been and untreated patients for either relapse-free survival or recruited. Antigenics has also filed with the European overall survival. Disturbingly, though, the independent review Agency (EMEA) for approval in Europe. The EMEA policy of committee also determined that 40% of the patients originally granting conditional approval would allow Oncophage to be logged by principal investigators in the multicenter trial as launched commercially in Europe provided Antigenics commits having had recurrences, in reality had residual disease before to conducting (continued)

nature biotechnology volume 27 number 2 FEBRUARY 2009 137 feature

Box 3 Personalized vaccines—a viable option? (continued)

a full-sized confirmatory phase 3 trial. Such an option is not This summer, Biovest reported that BiovaxID treatment available in the United States. prolonged disease-free survival by over one year, from 30.6 Like Antigenics’s Oncophage, BiovaxID, developed by Biovest months for control subjects to 44.2 months for BiovaxID-treated International, (Worcester, MA, USA) a majority-owned subsidiary subjects, (P = 0.047). The announced results will be formally of Accentia Biopharma (Tampa, FL, USA), is personalized and presented at the next American Society of Clinical Oncology targets antigens that are both tumor specific and unique to each Conference in Orlando and will be submitted for peer review. In patient. But its construct is entirely different. The vaccine’s addition to seeking approval in the United States, the company lead indication is indolent follicular non-Hodgkin’s lymphoma, is approaching regulatory agencies in other countries and plans in which a particular cancerous clone of antibody-producing on launching a compassionate-use program, referred to as B-lymphocytes proliferates. Name-Patient Program, for BiovaxID, in parts of Europe early in The BiovaxID concept began in the Stanford University laboratory 2009, according to Stergiou.. of Ron Levy and was pushed forward by Larry Kwak, who took The apparent BiovaxID success follows failures of two other the idea with him to the National Cancer Institute (NCI), where a nearly identical candidates, advanced by Genitope (Fremont, phase 2 trial was initiated in 1995. Kwak now heads the lymphoma CA, USA) and Favrille (San Diego), to reach statistically division at MD Anderson and consults for Biovest. significant results in phase 3 trials over the past year. Like All the constituent cells of a B-cell lymphoma produce BiovaxID, these were anti-idiotype vaccines conjugated to antibodies with identical idiotypes characteristic of the cancerous KLH and delivered with GM-CSF, but Stergiou speculates cells’ hyperproliferative common ancestor. Those malignant that the method of preparation of the antibody may be the B-cells also carry the idiotype on their surfaces. BiovaxID is an difference. Both Genitope and Favrille produced their antibodies anti-idiotype vaccine consisting of hybridoma-produced identical as recombinant proteins rather than using the hybridoma copies of those overabundant (and, therefore, easily characterized) methodology employed by Biovest. antibodies conjugated to the immune stimulant KLH. The vaccine Another potentially big difference is that in the BiovaxID trial, is administered along with GM-CSF. all vaccinated patients were in a state of complete remission. In 1999, favorable phase 2 results led the NCI to initiate The other candidates’ protocols, in contrast, allowed patients in a double-blind phase 3 trial, which Biovest took over under a partial remission—with visible tumor masses—to remain under Cooperative Research and Development Agreement. In the trial, study. Although this sped enrollment, it may have hindered 76 patients in remission following standard chemotherapeutic efficacy. regimens, received serial BiovaxID injections. Importantly, BiovaxID’s approach could, in theory, be applicable to other only subjects who had sustained a complete remission after non-Hodgkin’s lymphomas as well as multiple myeloma. But, chemotherapy were included, because previous studies showed that as a personalized vaccine treatment that must be produced patients in complete remission mount a better immune response batch by batch for each patient, even the most efficiently (humoral and/or cellular) compared to those who do not, which also manufactured product is likely to be expensive for patients— correlates with clinical outcome, according to Angelos Stergiou, although Stergiou refuses to put a price tag on the vaccine at chief medical officer and head of clinical research at Biovest. the moment.

propositions. Putting a novel approach through in trials, calls early attempts at vaccines tolerance and the hostile microenvironment in its paces among those most likely to benefit— “immunological kindergarten.” In addition, tumors—are combination therapies, but this patients who are least ill and for whom obtain- certain assumptions about the immune system creates certain problems, according to Robert ing statistically significant results will thus may not be correct. Early failures with cancer Schreiber, cancer researcher at Washington presumably take the most patience—is a costly vaccines led to the belief that tumors could not University School of (St. Louis). venture. Even the most sponsor-friendly phase generate an immune response, according to Eli “Most companies are locked into using their 3 cancer immunotherapy trials—a modest Gilboa, at the University of Miami, Florida. In own products and therefore do not like to use 300-patient study of an easily manufactured, fact, he says, “It appears [tumors] can [gener- combination therapies. And the FDA is par- mass-produced off-the-shelf vaccine, in an ate an immune response] for a time, but they ticularly leery of trying too many combina- indication with fast clinical readouts—and have elaborated mechanisms for avoiding the tions at once,” he says. Schreiber sees a role associated surgeries, imaging assays and so forth immune system. Focusing more on how to for university-industry partnerships in get- are going to cost about $20 million, according mitigate tumor-induced immune suppression ting around this potential logjam. “Once a to one knowledgeable company official. With will be key going forward.” successful regimen has been identified, there more and larger trials of personalized, more A second assumption that is proving false is will be many companies that come knocking technology-intensive vaccines, the cost soars that chemotherapy and immunotherapy are at the door. Since large-size clinical trials are to hundreds of millions. incompatible. According to Gilboa, evidence very expensive, I see a great opportunity for is emerging that some forms of chemotherapy industry and academia/foundations to pair up The way forward are not incompatible, but in fact can synergize at this time,” he says. All segments of the sector—immunologists, immunotherapy. As for cell-based therapies, the jury is still entrepreneurs, even regulators—appear to Another area of agreement is that we have out on whether autologous vaccines will be the agree that the entrée into cancer vaccines was reached the end of the single-agent era. Key to ticket or whether there is a place for allogeneic, premature. Thomas Okarma, CEO of Geron confronting the two main issues facing vaccine off-the-shelf ones. Logistics (read ‘cost’) seems (Menlo Park, CA, USA), which has a product developers—the ability of tumors to induce to dictate that only allogeneic vaccines will be

138 volume 27 number 2 FEBRUARY 2009 nature biotechnology feature

Table 5 Approved and marketed cancer vaccines Company (location) Product Description Indication Status Antigenics OncoPhage Heat shock protein vaccine puri- Renal cell carcinoma Approved in Russia fied from autologous tumor cells Granted fast track status by US FDA Biovest International BiovaxID Tumor-specific idiotype conju- Various B-cell–related cancers Compassionate use in France, gated to keyhole limpet hemocya- Germany, Italy, Greece, Spain nin, plus GM-CSF and the UK. Granted fast track status by US FDA Corixa Melacrine Lysate from two melanoma cell Melanoma Approved in Canada (acquired by GSK in 2005) lines, Detox adjuvant (proprietary) with monophosphoryl lipid A and mycobacterial cell wall skeleton CreaGene CreaVaxRCC Autologous monocytes treated Metastatic renal cell carcinoma Approved in Korea (Seoul) with GM-CSF and IL-4 to create immature dendritic cells acti- vated with tumor extracts plus cytokines Genoa Biotechnologia Hybricell Autologous monocytes treated Various cancers Approved in Brazil (Brazil) with cytokines and converted to dendritic cells that are fused with patient-derived tumor cells Vaccinogen OncoVax Metabolically active, irradiated, Colon cancer Approved in Europe, available (Frederick, MD, USA) autologous tumor cells with BCG in Switzerland Granted Fast Track status by FDA Mologen dSlim/Midge Genetically modified allogeneic Kidney cancer Orphan drug status granted (Berlin) (human) tumor cells for the by EMEA in 2006 expression of IL-7, GM-CSF, CD80 and CD154, in fixed combination with a DNA-based double stem loop immunomodulator (dSLIM). Center of Molecular Immunology CimaVax EGF EGF conjugated to rP64k Lung cancer Cuba, Peru (Cuba)

90 Vaccine candidates tools they need to create off-the-shelf vaccines. 80 For example, Geron, which has an autologous Therapeutic candidates 70 vaccine in trials now, is using this as a proof of 60 concept according to Okarma. The company 50 also has in place the technology for making 40 dendritic cells from stem cells, which would enable the company to prepare an off-the- 30 shelf, activated dendritic cell, something not Number of products 20 available at present. 10 Keith Wonnacott, chief of FDA’s cell thera- 0 pies, joins the chorus of immunologists and 6 8 0 3 academics optimistic that some cancer vac- 1990 1991 1992 1993 1994 1995 199 1997 199 1999 200 2001 2002 200 2004 2005 2006 cine will succeed. “We anticipate success, and Year that lessons will be learned. Much of what Figure 4 New cancer therapeutics and vaccines entering clinical study per year from 1990 to 2006. has gone on has been helpful. We would love (Source: Tufts Center for the Study of Drug Development) to see success,” he says. Whether that opti- mism is justified, only time will tell. commercially viable, although the evidence immunogenicities? Different cross-reactivi- to date suggests that it may not be a clinically ties? Different potencies? 1. Choudhury, A. et al. Adv. Cancer Res. 95, 147–202 (2006). viable approach. The answer might rest in finding more 2. Valone, FH, et al. Cancer J., 7, S53–61 (2001). Even non-cell-based personalized vaccines shared, but tumor-specific, antigens, which are 3. Small, E.J. et al. J. Clin. Oncol. 18, 3894–3903 raise a host of regulatory issues. Is each vac- in the minority among products in trials today. (2000). 4. Allison, M. Nat. Biotechnol. 26, 967–969 (2008). cine a different product? Do vaccines pro- And down the road, advances in related fields 5. Testori, A. et al. J. Clin. Oncol. 26, 955–962 (2008). duced for different patients have different might provide cancer vaccinologists with the 6. Wood, C. et al. Lancet 372, 145–154 (2008).

Corrected after print 7 June 2010.

nature biotechnology volume 27 number 2 FEBRUARY 2009 139 errata

Erratum: The cancer vaccine roller coaster Bruce Goldman & Laura DeFrancesco Nat. Biotechnol. 27, 129–139 (2009); published online 7 February 2009; corrected after print 7 June 2010

In the version of this article initially published, the Mologen product description in Table 5, page 139, was incomplete and its status incorrectly stated to be compassionate use in India. The product description should have read: Genetically modified allogeneic (human) tumor cells for the expression of IL-7, GM-CSF, CD80 and CD154, in fixed combination with a DNA-based double stem loop immunomodulator (dSLIM). The status should have read: Orphan drug status granted by EMEA in 2006. The error has been corrected in the HTML and PDF versions of the article.

Erratum: Irish bioethics council axed Cormac Sheridan Nat. Biotechnol. 28, 112 (2010); published online 5 February 2010; corrected after print 7 June 2010

In the version of this article initially published, a researcher at University College Cork was incorrectly named. His name is Tom (not Barry) Moore. The error has been corrected in the HTML and PDF versions of the article.

Erratum: Never again Chris Scott Nat. Biotechnol. 28, 131 (2010); published online 5 February 2010; corrected after print 7 June 2010

In the version of this article initially published, Art Levinson is incorrectly described as a founder of Genentech, Sandra Horning as senior vice president of global clinical development and Richard Scheller as chief of operations. Their titles should have read: CEO Arthur Levinson moved up to the board of directors…. Sandra Horning…took over as senior vice president, global head, clinical development, hematology/oncology. Executive vice president, research, Richard Scheller…. The errors have been corrected in the HTML and PDF versions of the article.

Erratum: Resuscitated deCODE refocuses on diagnostics Mark Ratner Nat. Biotechnol. 28, 192 (2010); published online 8 March 2010; corrected after print 7 June 2010

In the version of this article initially published, it was reported that deCODE had “shuttered its Emerald Biosciences and Emerald Biostructures drug

© 2010 Nature America, Inc. All rights reserved. All rights Inc. America, Nature © 2010 discovery operations”; in fact, the companies were sold to investors. In addition, the correct name of Emerald Biosciences is Emerald BioSystems. The error has been corrected in the HTML and PDF versions of the article.

Erratum: Biotech in a blink Ken Garber Nat. Biotechnol. 28, 311–314 (2010); published online 8 April 2010; corrected after print 15 April 2010

In the version of the article originally published, Michael Tolentino was misquoted to the effect that bevasiranib had been shown to persist indefi- nitely in post-mitotic cells. Tolentino actually stated that the RNA-induced signaling complex persists. The error has been corrected in the HMTL and PDF versions of the article.

624 volume 28 number 6 JUNE 2010 nature biotechnology