Immunotherapy for Lung Cancer Edward A. Hirschowitz1,2 and John R. Yannelli3 1 Department of Internal Medicine, Division of Pulmonary and Critical Care Medicine, University of Kentucky, Chandler Medical Center, Lexington, Kentucky; 2Lexington Veterans Administration Medical Center, Lexington, Kentucky; and 3Department of Microbiology, Immunology, and Human Genetics, University of Kentucky, Lexington, Kentucky Immunotherapy is a conceptually attractive approach, because it is trastuzumab (Herceptin), which target epidermal growth factor highly speciﬁc and can deal with disseminated disease with minimal (EGF) receptors, HER-1 and HER-2, respectively, and bev- impact on normal tissues. Ability to induce antigen-speciﬁc immune acizumab (Avastin), which interferes with tumor angiogenesis responses in patients with lung cancer is now well established in by binding to vascular endothelial growth factor (1–4). Adop- early-phase clinical trials using a variety of immunotherapeutic tive transfer is another example of passive immunotherapy that approaches. Although no immunotherapy is likely to be a panacea, typically involves ex vivo ampliﬁcation and infusion of autolo- randomized phase IIB studies offer promise of therapeutic activity in gous tumor-inﬁltrating T cell or lymphokine-activated killer cell both early- and late-stage lung cancer. This review will cover basic therapy (5). By contrast, active immunotherapy uses the host’s concepts of immunotherapy, provide perspective on vaccine devel- immune cells and requires an intact immune system to function. opment, and update the status of ongoing clinical trials in lung Active immunotherapy is derived from the knowledge that the cancer. immune system can discriminate cancer cells from normal cells Keywords: immunotherapy; vaccines; lung cancer; clinical trials based on tumor antigen recognition (6–9). Approaches that directly incorporate tumor antigen are conventionally referred to as vaccines. Although some active immunotherapies are INTRODUCTION designed to induce antibodies as the primary effector mecha- nism, and may be rationally applied with certain logical in- A number of nonconventional therapeutic modalities are being tention, cancer-speciﬁc antibody responses are widely viewed as developed to improve unacceptably poor outcomes in lung having limited direct cytotoxic capability against cancer. With cancer. Immunotherapy is an attractive approach that, by few exceptions, induction of antigen-speciﬁc T-cell responses is design, is cancer speciﬁc and can target disseminated disease the primary objective of active immunotherapy (6–10). Natural with minimal impact on normal tissues. Active immunotherapy killer (NK) cells are an antigen-independent arm of immune describes those approaches that use host immune machinery to defense that targets major histocompatibility complex (MHC)- generate antitumor effects. This review presents basic concepts mismatched (allogeneic) cells or those cells lacking surface of active immunotherapy, provides perspective on therapeutic MHC molecules. Because cancer cells often downregulate development, and updates the status of ongoing clinical immu- surface MHC molecules, induction of NK is a logical, but not notherapy trials in lung cancer. common, primary objective of immunotherapy (11). IMMUNOTHERAPY AND IMMUNOBIOLOGY Immune Recognition Immunotherapy Identiﬁcation of tumor-associated antibodies and antigen- speciﬁc cytotoxic T lymphocytes (CTL) in patients with lung Novel therapies should improve outcomes, but have a side- cancer indicates that the immune system can distinguish lung effect proﬁle that is more favorable than conventional chemo- cancer cells from normal cells (6, 7). Antigen-speciﬁc immune therapy (1). A number of therapeutic approaches are being responses are dependent on antigen presentation in the patient’s developed that selectively target malignant cells or their lymphoid tissues. When antigen is contacted, antigen-presenting microenvironment, but leave normal cells intact. An ideal agent cells (APCs) digest whole proteins into smaller peptides that are should also be easy to administer to address both quality-of-life then presented on HLA class I or class II molecules on the APC issues and compliance concerns. Although the term ‘‘targeted surface. Coordinated expression of costimulatory molecules therapy’’ has been most widely applied to pharmacologic (CB80/CD86) on the APC surface prompts antigen-speciﬁc agents, including small molecule inhibitors and antisense oligo- lymphoid precursors to activate at appropriate times. Depending nucleotide, immunotherapy closely ﬁts the deﬁnition. on the type or source of antigen and the existing cytokine milieu, Immunotherapy is categorized as either passive or active. class II peptides may stimulate IL-2 and IFN-g release by antigen- Passive immunotherapy includes any immunologically active speciﬁc CD41 T cells (Th1). Antigen-speciﬁc CTL are generated agent that is made outside the body and does not rely on host when CD81 T cells bound to class I antigens on APCs are machinery to function. The most widely applied passive immu- stimulated by Th1 cytokines. Alternatively, class II antigen notherapies are monoclonal antibodies that disrupt tumorigenic binding may induce a different T-cell phenotype (Th2) that cascades by blocking the binding of hormones or growth factors releases IL-4 and IL-10, and interacts with B cells to promote to their receptors; examples include cetuximab (Erbitux) and antigen speciﬁc antibody production. While Th1 and Th2 are not necessarily mutually exclusive in the immune response, Th2- related cytokines tend to suppress Th1-mediated responses. (Received in original form June 16, 2008; accepted in ﬁnal form July 21, 2008) Notably, activated Th cells have longevity compared with Correspondence and requests for reprints should be addressed to Edward A. other activated immune cells (APCs and CTL), and are conven- Hirschowitz, M.D., Division of Pulmonary and Critical Care Medicine, University of tionally viewed as being responsible for ‘‘immune memory,’’ Kentucky, Chandler Medical Center, 740 S. Limestone, Room K528, Lexington, capable of ‘‘revitalizing’’ the immune response if a speciﬁc KY. E-mail: firstname.lastname@example.org antigen is reencountered; memory subsets of CD81 T cells and Proc Am Thorac Soc Vol 6. pp 224–232, 2009 DOI: 10.1513/pats.200806-048LC B cells provide other mechanisms for rapid response to known Internet address: www.atsjournals.org antigens. Hirschowitz and Yannelli: Immunotherapy for Lung Cancer 225 Immune Regulation and Tumor Evasion and confounding comparability within the study population, Physiologic checks and balances, that work throughout the both of which can hinder rational therapeutic development in immune cascade to control antigen hyperresponsiveness, create early clinical testing. Nonetheless, autologous vaccines have a permissive environment for tumor growth (9, 11, 12). Immu- been shown to have immunologic activity in a number of studies nosuppressive cytokines, APC senescence, and regulatory T (10, 11). Allogeneic tumor is a rational alternative that can cells (T-regs) each present an important challenge to successful provide a renewable source of antigen and offers greater development of immunotherapy (9, 11–13). Most established potential for ‘‘off-the-shelf’’ application. Using allogeneic anti- tumors also have mechanisms of immune evasion (9, 11, 12). gen also generates a uniform preparation, which facilitates The production and induction of immunosuppressive cytokines immune assessment and comparability not afforded by use of by tumor cells, such as IL-10 and transforming growth factor-b, autologous tumor antigen. Availability of these tangible mile- has been well characterized (12). Some tumor cells may avoid stones makes this, and other allogeneic approaches, attractive immune destruction by down-regulating HLA class I molecules during therapeutic development and clinical testing. on which antigens are presented for recognition, or by over- Two additional allogeneic sources of antigen are synthetic expressing B7-H1, a ligand for the T-cell receptor, PD-1, which peptide and recombinant protein (including gene therapy– is known to negatively regulate T-cell activation. Deﬁned derived antigen). In contrast to allogeneic tumor, peptides and mechanisms of CTL resistance and counter defense include proteins are characteristically used in monovalent formulations. tumor expression of the apoptosis-inducing molecule, Fas Peptides are an attractive monovalent antigen source, because ligand, and downregulation of surface Fas receptors (12). they are small and easily synthesized, are uniform, and provide the simplest and most reproducible immunologic measures of biological efﬁcacy (11). A disadvantage is that peptides are CANCER ANTIGENS exclusive to speciﬁc HLA types and require patient selection There are multiple incarnations of antigen used in active based on HLA tissue typing. Furthermore, peptides have de- immunotherapy, each with independent advantages. A major signated restriction to class I or class II pathways, selectively division in antigen targeting is the use of a multivalent or stimulating either CD81 cytotoxic T-cell effectors or CD41 Th monovalent strategy. By targeting several antigens, multivalent cells, responsible for immune memory (11). By contrast, whole approaches lend themselves to many of the clinical and bio- recombinant proteins are processed into multiple peptides and logical realities of lung cancer. Speciﬁcally, phenotypic hetero- presented by APCs via class I and class II pathways to CD41 and geneity characteristic of non–small cell lung cancer (NSCLC) CD81 T cells, respectively, and have the potential for generating and small cell lung cancer (SCLC) dictates high variability of a response that includes immune effectors and immune memory. antigen expression, and precludes the ideal circumstance of one Xenogeneic antiidiotype antibodies are somewhat unique antigen being uniformly expressed by all lung tumors. Addi- antigen-mimic preparations, generated as antibodies to tumor tionally, because a vast majority of patients with lung cancer antigen–binding sites on other antibodies (that generates a tem- are nonsurgical, most lung cancer diagnoses are made with plate of the antigen). The foreign (xenogeneic) nature of these minimal tissue sampling, and the material is often too limited preparations makes them inherently immunogenic, and the for adequate antigen characterization; in context, multivalent similarity of the antiidiotype antibody to the tumor antigen approaches circumvent the need to know which antigens are allows cross recognition of the parent/native protein (14). In expressed by a speciﬁc tumor. By contrast, monovalent applica- contrast to the vast majority of the other active immunothera- tions all have the common feature of being selectively delivered to pies under development, antiidiotypic vaccines are used to elicit individuals with corresponding tumor antigen expression, which tumor-speciﬁc antibodies as the dominant effectors for thera- distinctly requires individual tumor antigen characterization. peutic activity; these have been the most widely tested immu- Notably, monovalent approaches are more readily incorporated notherapy approaches in SCLC. into proprietary products, and lend themselves to commercial development; not surprisingly, a majority of therapies advancing PROMOTING ANTIGEN RECOGNITION: IMMUNE to phase III investigation are monovalent approaches. ADJUVANTS AND DELIVERY The most commonly used multivalent formulations employ autologous or allogeneic tumor cells, although multiprotein/- Success of any active immunotherapy ultimately depends on peptide mixtures or fusion constructs can achieve similar tumor antigen capture and presentation by APCs (11, 15). multivalent objectives. Tumor-derived antigen mixtures include Tumor antigens, however, are not, in themselves, immunogenic. multiple dominant and minor antigenic determinants within In order to initiate or promote antigen-speciﬁc responses, all whole proteins, allowing the host to select, process, and present strategies incorporate adjuvants; these are essentially activating on MHC, the most immunogenic epitopes relevant to that agents or stimulants for various arms of the immune cascade, individual (11). Autologous tumor can be utilized either intended to augment antigen recognition, uptake, presentation, in vivo or ex vivo. The former relies on a locoregional or and/or antigen-speciﬁc cellular reactivity (11, 15). Effects of systemic intervention to promote antigen recognition in situ. various adjuvants are not mutually exclusive, and mechanisms One example of this is the systemic administration of cytokine; often overlap and intersect. There are myriad choices that fall alternatively, if the tumor bed can be directly accessed, gene into one of several categories: (1) biologic and chemical therapy strategies can be used to modify tumor cell antigenicity adjuvants; (2) recombinant cytokines and chemokines; (3) auto- or alter the tumor microenvironment. Ex vivo approaches most logous dendritic cells (DCs); (4) immune modiﬁers; (5) gene commonly combine surgically obtained tumor with an immune therapy/gene transfer vectors. adjuvant to produce a cancer vaccine. A disadvantage of using autologous tumor in vaccines is that adequate amounts of fresh Chemical and Biologic and Adjuvants tumor must be available for vaccine production, generally These adjuvants have been used for decades as a key compo- restricting this approach to the individual whose tumor is nent in antigen-speciﬁc immunotherapy to induce inﬂamma- surgically resected with the forethought and intent of making tion, either at the site of tumor, or used in conjunction with a vaccine. An additional limitation is that antigens differ for exogenously supplied antigen. Biologic adjuvants take advantage each subject, thereby complicating immunological assessment of the fact that they are immunogenic compounds, naturally 226 PROCEEDINGS OF THE AMERICAN THORACIC SOCIETY VOL 6 2009 recognized as foreign and known to induce migration of APCs to to cytotoxic therapy, where other effects may be unique to the site of delivery. APCs responding to adjuvant stimulation are speciﬁc chemotherapeutic classes or agents. It has been postu- thus able to coincidentally capture and process tumor antigens lated that apoptosis of tumor cells induced by most commonly present in the inﬂammatory milieu. The most commonly used used cytotoxic therapies somewhat paradoxically stimulates the ´ biologic adjuvants include bacillus Calmette-Guerin (BCG), immune system (25–27). Therapy may also mitigate inhibitory diphtheria toxoid, and tetanus toxoid (11, 15). Chemical adju- cytokine production in the local milieu, or induce proinﬂamma- vants function similarly to biologic adjuvants as irritants that tory signals that augment APC numbers and function or promote induce an inﬂammatory response at the site of delivery, and some generation of antigen-speciﬁc CTL (25, 26, 28). The literature may also provide a matrix that sequesters antigen at a speciﬁc also describes several drug-related phenotypic alterations that location, allowing a timed release of antigen to APCs. Examples lead to increased susceptibility to CTL lysis, including the in- include aluminum hydroxide, montanide ISA 51, and incomplete duction of tumor antigens and chaperone molecules (heat shock Freund’s adjuvant (11, 15, 16). proteins), upregulation of MHC, intercellular adhesion molecule, tumor necrosis factor–related apoptosis-inducing ligand, and Fas Recombinant Cytokines and Chemokines receptors, each of which could facilitate immune-mediated Systemically administered cytokines have also been explored recognition and destruction of tumor cells (22, 26, 29–33). for decades as stand-alone agents to nonspeciﬁcally activate ‘‘Antisuppressive agents’’ are a subcategory of immune mod- CTL (and/or NK cells), to skew the immune response toward ulators with capacity to reduce regulatory elements in the host a CTL response through effects on Th cells and APCs, to induce environment—most speciﬁcally, T-regs. Abrogation of T-regs APC differentiation and function, or to promote HLA class I has been most extensively described with cyclophosphamide and, molecule expression on tumor cells. Cytokines have also been to a lesser extent, with the antimetabolite, ﬂudarabine (22–26, combined with vaccines, most notably by using cytokine–gene 34). Ongoing development and testing of selective inhibitors of transfer to generate high local concentrations of stimulatory T-regs is likely to yield multiple agents with therapeutic potential cytokines at the site of antigen delivery (11, 17). Granulocyte- that could synergize with cancer vaccines and other active macrophage colony–stimulating factor (GM-CSF) and IL-2 have immunotherapy strategies (11, 25). been the two most widely investigated cytokines in clinical trials. The list of cytokines is far too extensive to be detail here, and the Gene Therapy and Gene Transfer Vectors reader is directed to References 11 and 17 for a further discussion Gene therapy is not a uniquely independent category, but, of therapeutic cytokine uses in cancer. Similar to cytokines, rather, describes a strategy that employs a transfer vector for chemokines may be used at the site of tumor or combined with direct expression of gene-derived proteins that modify cells for exogenous antigen to attract APCs to a site of vaccination. alternate function. Because the designation of gene therapy as a ‘‘therapy’’ can be misleading, an appropriate distinction Autologous DCs should be made between gene transfer vectors and gene-based DCs are professional APCs that possess all necessary elements therapeutic proteins (oncogenic viruses are a notable exception, to initiate and potentiate an antigen-speciﬁc immune response. as the innate cytolytic properties of these viruses are being Most commonly, DC precursors are harvested by the leukaphe- employed therapeutically). A variety of gene transfer vectors, resis procedure from a patient with disease, cultured in vitro, each with differing rates of efﬁciency, can be coupled with and supplied antigen ex vivo. When delivered back to the innumerable gene-derived proteins with adjuvant properties. individual, it is expected that the programmed DC migrates to Gene transfer of cytokines or costimulatory molecules di- lymph node in vivo and orchestrates the desired antigen-speciﬁc rectly to tumor cells ex vivo and in vivo is an attractive way of immune response (18). making nonimmunogenic cells more immune stimulatory (11). In vivo cytokine gene transfer can also target normal cells in the Immune Modulators tumor milieu, thereby achieving high local concentrations of cytokine that avoid toxicities associated with systemic adminis- Most immune modulators are pharmacologic agents that have tration. Other gene-based vaccine strategies modify normal independent therapeutic effects, but also have immunologic cells in vivo to express and present tumor antigens; dual properties that promote a favorable immune environment. expression of antigen and gene-based cytokine or costimulatory Examples include cyclooxygenase-2 inhibitors and thalidomide- molecule expression has also been explored as a means of like agents (called ‘‘immunomodulatory drugs’’ or IMiDs), such enhancing biological efﬁcacy. Secondary advantages of viral as lenalidomide (Revlimid). These and other similar agents have gene-transfer vectors used in vivo or ex vivo is an adjuvant the signiﬁcant potential for synergy with cancer vaccines (19, 20). stimulation of the immune system responding to viral proteins. Small molecules that stimulate Toll-like receptors, a class of This may be an important factor in experimental success with molecular mediators involved in the initiation of the innate and gene-based antigen vaccines (35). adaptive immune responses, have also been shown to activate DCs and induce a favorable cytokine environment. CpG oligodeoxynucleotides are Toll-like receptor-9 agonists being CLINICAL APPLICATION evaluated for monotherapy in clinical trials, but also hold promise as adjuvants for cancer vaccines (21). Clinical Population Interestingly, there is an emerging body of literature that Based on poor treatment outcomes in advanced-stage lung suggests that chemotherapy, conventionally viewed as antago- cancer, investigating therapies that may consolidate clinical nistic with immunotherapy, can have adjuvant properties that responses to chemotherapy and radiation is warranted. Simi- somewhat counter-intuitively promote immunological objec- larly, adjuvant therapies that could destroy limited residual tives (22–26). Beyond the relative clinical and biological impor- disease and small metastatic deposits after surgical resection are tance of tumor debulking, cytotoxic therapies have numerous highly attractive. The low risk and potential beneﬁt of tumor systemic and local effects that might lead to synergy of immuno- vaccines is appealing in both populations, but is most pro- and chemotherapy; it can be hypothesized that multiple mech- nounced for patients having surgically resected NSCLC; mini- anisms act in concert. Some described mechanisms are generic mal tumor burden is generally perceived as the most amenable Hirschowitz and Yannelli: Immunotherapy for Lung Cancer 227 clinical target for immunotherapy, and the nominal risk of the during therapeutic development. To date, there is not a standard intervention is highly relevant, because we cannot yet predict assay nor consensus on what constitutes a positive immunologic recurrence beyond statistical probability. In contrast, patients response (36). Considerations when choosing assays for analysis with advanced-stage lung cancer, especially those individuals include the information desired (induction of antibody, T cell, with bulky disease, have not been widely viewed as likely or T-cell subsets), the number of samples to be assayed, and the to beneﬁt from immunotherapy. Beyond mechanical factors volume of sample obtainable from each subject on serial blood related to tumor burden, there are numerous biological ele- draws. Serum antibody measurement is well established, stan- ments that could negate potential therapeutic effects of immune dardized, simple, and accurate; nonetheless, few vaccines are intervention. Notably, larger and metastatic tumors are known designed to generate antibodies as a primary effector mecha- to acquire enhanced resistance capability during progression; nism, and the therapeutics relevance of antibody is uncertain. immunosuppressive elements are also more prominent in Conversely, measurement of antigen-speciﬁc T cells in samples advanced-stage cancer, including a purported corresponding from immunized patients is less-well standardized, laborious, increase in T-regs with increasing tumor burden. Nonetheless, but more highly relevant to desired effects of most immuno- multiple studies indicate that vaccines can induce immune therapies. responses in a percentage of patients with advanced-stage Functional T-cell assays include cytotoxicty, proliferation, disease, and data from completed phase IIB clinical trials and cytokine production assays that all measure T-cell reactivity suggest that lung cancer vaccines may slow disease progression upon antigen stimulation. Each is employed to demonstrate higher and/or improve quality of life in this population. When assessing frequency of antigen-speciﬁc T cells in circulating peripheral blood efﬁcacy of immunotherapy for advanced-stage NSCLC, it is mononuclear cells after vaccination compared with baseline important to keep in mind that an added therapy that may only (prevaccine). Interpretation is based on the notable assumption provide marginal improvements in survival or quality of life, but that the T-cell responses of peripheral blood mononuclear cells has minimal risk, is commensurate with the deﬁnition of beneﬁ- reﬂect responses in draining nodes or site of tumor, and analysis cial treatment currently applied to conventional chemotherapy. may over- or underrepresent compartmental activity. Delayed- Immunotherapy for SCLC has not been as widely investi- type hypersensitivity to antigen challenge is a qualitative func- gated. The disease tends to be very aggressive, and immuno- tional assay performed in vivo. Similar to the Mantoux skin test for therapy has not been widely regarded as having a signiﬁcant tuberculosis, a mononuclear cell response is mounted at the site potential to impact outcomes. Because SCLC is an exclusively of antigen challenge if the patient has preexisting T-cell immunity. nonsurgical disease, tumor is not routinely available for autol- This is a widely used assay that can be practically applied to most ogous vaccine production. The lack of pathologic specimens has vaccine approaches; however, sensitivity is limited and results in also slowed preclinical investigation, as SCLC is not available clinical trials have been variable. Preexisting immunity, which may for in vitro study or xenogeneic tumor modeling, nor does there be induced by a patient’s own tumor, and biological differences exist an appropriate SCLC animal model. Regardless, the poor in immune competence can cloud the readout. Tetramer staining prognosis associated with SCLC makes it a rational choice for is a nonfunctional, quantitative measure of antigen-speciﬁc T-cell further investigation, and several groups are auspiciously pur- frequency in peripheral blood. Florescent-labeled tetramers are suing vaccines for SCLC. constructed to bind a unique MHC/peptide-speciﬁc T-cell re- ceptor, and tetramer-tagged T cells are quantiﬁed by ﬂow cytom- Clinical Testing etry. MHC class I–peptide tetramers used to measure CD81 CTL speciﬁc for select antigens are the most commonly available, The fundamental objective of clinical immunotherapy testing is although some MHC class II–peptide tetramers have also been to gain relevant biological and/or therapeutic information that developed for the assessment of CD41 T-cell responses. Because promotes rational development of therapeutic strategies. The tetramers are speciﬁc for a single, speciﬁc HLA-matched epitope, complexity of the immune system and reliance on an incom- this assay is most practical for use with HLA-restricted peptide pletely characterized series of events makes this a challenging vaccines. Tetramers are, however, not available for all antigens or prospect. Animal studies do provide insight into immunobiology, all HLA types (the reader is directed to References 36 and 37 for and have been critical in rational design of immunotherapy comprehensive discussion of tools for immunologic monitoring of strategies; the parallels to human disease are, however, imperfect, cancer vaccine trials). and clinical trials are built on only partial understanding of human There are also a number of ancillary measurements that may cancer immunobiology. Although clinical study design is based reﬂect the receptiveness of the host to immune induction. on a fair amount of speculation, extrapolation from measurable Serum and cellular cytokine characterization has been used immunological parameters in early-phase clinical trials provides to gauge immune hyporesponsiveness, and to measure the effec- valuable information. A number of in vitro assays that measure tiveness of immune modulators. Similarly, immunological charac- a variety of parameters in peripheral blood after intervention terization of the tumor compartment has indicated presence of allow comparative appraisal of immunological activity. Impor- multiple elements that inhibit antigen presentation and neutral- tantly, observed correlation in some clinical trials between ize effector cells; these measurements are, however, neither immunologic activity (measured by a variety of assays) and readily available nor routinely sought. In parallel with evolving clinical response criteria strongly supports the hypothesis that knowledge of T-regs, measurement of percent circulating T-regs immunological response to relevant antigens translates into by cell surface cytometric analysis appears to offer an additional clinical beneﬁt. Deﬁnitive evidence of therapeutic effect in phase gauge of host immune responsiveness. CD41CD251FoxP31 III trials and an established link between immunological assays T-regs are most well characterized, although several different and clinical outcomes will both facilitate and promote logical T-reg populations have been identiﬁed. Notably, the literature development of cancer immunotherapy. reports variably elevated levels of CD41CD251FoxP31 T-regs in peripheral blood and in tumor beds of patients with lung cancer, Immunologic Monitoring a ﬁnding that may correlate with stage and/or prognosis (38–43). Immunologic endpoints are critical for determining biologic T-reg measurements may prove to be a relative predictor of activity of active immunotherapies in phase I and phase II resistance to immune induction and become a directed endpoint studies, and serve as reasonable measures of vaccine potency for immune modulation. 228 PROCEEDINGS OF THE AMERICAN THORACIC SOCIETY VOL 6 2009 Clinical Endpoints strongly support a correlation between clinical and immunolog- Clinical endpoints are the foundation of phase IIB and phase III ical response. trials. Rate of recurrence or risk of recurrence is the primary clinical endpoint in studies with surgically resected patients with ADVANCED STUDIES USING SELECTED AGENTS lung cancer; advanced-stage disease offers several measures of efﬁcacy, including objective tumor response, duration of re- A handful of immunotherapies for lung cancer have been sponse, time to progression (TTP), disease-free survival, and investigated in advanced-phase studies. Nine vaccines tested overall survival. Objective tumor response may be measured in independent trials are discussed here (an overview is by Response Evaluation Criteria in Solid Tumors (RECIST) or presented in Table 1). All but one study has been conducted World Health Organization criteria, but are reserved for patients in NSCLC; the only SCLC vaccine of this group is also the only with measurable disease at baseline; this typically includes serial one to have been completely evaluated in a phase III trial. One measurement of one or more target lesions by computed tomog- other study in NSCLC, designed as a phase III, had a high raphy. Increasingly sensitive radiographic techniques that have dropout rate and has statistical power that more closely resem- yet to be validated for clinical investigation may provide newer bles a randomized phase IIB study. Four of the other seven clinical response criteria that could be used as intermediate agents discussed here have been tested in phase IIB, random- endpoints in phase II studies (44). Tumor markers may be used ized, controlled trials; the remaining three represent a series of as a supplementary measurement, but, alone, are not accepted as vaccines from a single group, which were each tested in larger, a measure of response. nonrandomized phase II studies. With few exceptions, adverse Although targeting early-stage disease is appealing, an events reported in all studies were limited to local site reactions, important consideration for therapeutic development is that and less frequently, ﬂu-like symptoms and fatigue. relatively low rates of recurrence in surgically resected patients dictates a large study sample size to show clinical beneﬁt in phase BEC2 and BCG III trials. By contrast, radiographically measurable disease, rapid BEC2 is an antiidiotypic vaccine for SCLC and is the only lung progression, and terminal prognosis of advanced-stage lung cancer vaccine to have been completely evaluated in phase III cancer provides timely endpoint analysis for therapeutic clinical studies. In the initial phase I/II trial, patients with limited-stage trials. The comparatively small study numbers required to see SCLC achieving a partial or complete response from chemo- statistical differences between study groups also makes this an therapy were vaccinated with an antiidiotype GD3 monoclonal attractive population for therapeutic investigation. Of note, antibody (BEC2) and BCG. Those who developed anti-GD3 novel molecular analysis may soon be able to determine which antibodies had prolonged survival compared with historical individuals are likely to recur after surgical resection of NSCLC, control subjects. These encouraging results led to a phase III making immunotherapy studies more rational and efﬁciently trial sponsored by Merck KGaA and ImClone. A total of 515 conducted in a reﬁned population of patients having had surgical patients were randomized 1:1 to receive ﬁve immunizations of resection. BEC2 plus BCG over a 10-week period or best supportive care (BSC); the phase III trial disappointingly did not conﬁrm clinical beneﬁt suggested by the early-phase study (45). Notably, NONRANDOMIZED CLINICAL TRIALS IN vaccine induced humoral response in only one-third of the 213 LUNG CANCER patients who were able to be evaluated immunologically (negative Myriad approaches with multiple antigens have been explored in 142 cases and positive in 71 cases). Stratiﬁed analysis suggested in a variety of malignancies. The published literature describes a trend toward improved survival in patients that showed immu- fewer than 600 patients with lung cancer that have been treated nological response, although differences were not statistically in 25 pilot, phase I, or early phase II studies using 17 different signiﬁcant. The authors concluded that the study was essentially vaccines. Categorization of these trials roughly parallels the negative; however, they acknowledged that the agent has limited classiﬁcation of immune adjuvants, including: (1) antigen plus ability to induce antibody response, and that the reduced number chemical/biologic adjuvant; (2) antigen plus cytokine; (3) anti- of patients in stratiﬁed groups likely limited the power of their idiotype plus biologic adjuvant; (4) antigen-loaded DCs; and (5) analysis. Furthermore, the authors conclude that a monovalent gene therapy. The objectives of each of these trials included approach may not have been ideal, indicating that multivalent safety, tolerability, and biological activity (10). approaches may be better (46). Comparison of these small studies is difﬁcult if not impos- sible (10, 36, 37). Each involves a different study population, Anti-EGF Vaccine vaccine protocols, doses, frequency, and even routes of admin- The anti-EGF vaccine is unique among other lung cancer istration. A number of different antigens have been incorpo- vaccines in that it does not directly target tumor, but, rather, rated into a variety of vaccine preparations, and responses were has the primary objective of inducing anti-EGF antibodies that assessed by variable means of immune assessment and non- neutralize endogenous EGF and deprive tumor of this important rigorous clinical response criteria. Nonetheless, the literature growth factor (47, 48). The vaccine is comprised of recombinant collectively indicates that immunotherapy for lung cancer is EGF, chemically conjugated to a recombinant P64K bacterial feasible and rational. Speciﬁcally, it is apparent that toxicity is protein as carrier protein, and emulsiﬁed with the adjuvant limited, and that multiple agents can induce measurable immu- Montanide ISA51. The randomized phase IIB study enrolled 80 nologic response. A consistent observation in all studies is that patients with stage IIIB/IV NSCLC who had completed ﬁrst- immune induction is not uniform, even within homogenous line chemotherapy. Patients were randomized 1:1 to BSC or study populations, indicating a biological variability that regu- EGF vaccinations. The treatment group received four induction lates individual responses, which is independent of vaccine immunizations and monthly boosters until disease progression. potency. Furthermore, although early-phase trials are not Vaccine induced a desirable anti-EGF antibody response in powered to deﬁne therapeutic efﬁcacy, several trials offer roughly 50% of immunized patients, which corresponded to a anecdotal evidence of clinical beneﬁt. An important and highly signiﬁcant decrease in serum EGF concentrations. There was a relevant corollary to these observations is that several studies signiﬁcant correlation between anti-EGF antibody titers, serum Hirschowitz and Yannelli: Immunotherapy for Lung Cancer 229 TABLE 1. RANDOMIZED CLINICAL VACCINE TRIALS IN LUNG CANCER: PHASE IIB AND PHASE III RANDOMIZED STUDIES Tumor/ Randomization/ Endpoint Vaccine Description Stage Patient n Trial Design (Signiﬁcance) Secondary Analysis Ref. No. BLP25 MUC-1 peptide 1 NSCLC IIIB/IV 171 1:1, vaccine vs. BSC Survival (NS) Trend toward increased (54) (Stimuvax) liposome survival in subgroup with stage IIIB disease. MAGE-A3 MAGE protein 6 NSCLC I/II 182 2:1, vaccine vs. BSC Recurrence (NS) Trend toward delay in (55, 56) (ASCI) AS02B (MAGE-A31) time to recurrence/ nonstatistical reduction in relative risk of cancer recurrence. TG4010 Dual gene therapy NSCLC IIIB/IV 148 1:1, vaccine 1 chemo Survival (pending) Interim results indicate a (57) vaccinia/MUC1/IL-2 vs. chemo alone nonstatistical increase in 6-mo survival, with statistical differences in clinical response (WHO criteria). EGF EGF protein 1 P64 NSCLC IIIB/IV 80 1:1. vaccine vs. BSC Survival (NS) Signiﬁcant survival (47, 48) carrier protein 1 advantage in subgroup montanide ISA51 with vaccine-induced anti-EGF Ab response who were ,60 yr old; signiﬁcant correlation between anti-EGF antibody titers, serum EGF levels, and clinical outcomes (Ab response in only 50%). *SRL172 Heat-killed M. vaccae NSCLC IIIB/IV 419* 1:1, vaccine 1 chemo Survival (NS) Signiﬁcant survival (49, 50) 1 chemotherapy vs. chemo alone advantage in subgroup with adenocarcinoma; improved quality of life overall. †BEC2 Antiidiotype GD3 1 Small cell 515 1:1, vaccine vs. BSC Survival (NS) Low rate of immune (45, 46) BCG (limited) induction (1/3); trend toward improved survival in subgroup with vaccine-induced anti-GD3 Ab response. ´ Deﬁnition of abbreviations: Ab 5 antibody; ASCI 5 antigen-speciﬁc cancer immunotherapy; BCG 5 bacillus Calmette-Guerin; BSC 5 best supportive care; EGF 5 epidermal growth factor; M. vaccae 5 Mycobacterium vaccae; MUC1 5 mucin-1; NS 5 not signiﬁcant; NSCLC 5 non–small cell lung cancer; WHO 5 World Health Organization. * SRL172: designed as phase III trial; fewer than 50% completed the prescribed number of immunizations. † BEC2: the only completed phase III trial in lung cancer. EGF levels, and clinical outcomes. The authors report an 223 vs. 225 days; P 5 0.65). Secondary analysis showed that overall trend toward increased survival in the treatment group patients with adenocarcinoma who completed the vaccine pro- compared with control subjects that reached statistical signiﬁ- tocol (n 5 45) versus matched control subjects receiving cance in a subgroup of patients less than 60 years old (47, 48). chemotherapy alone did have a survival advantage (median overall survival, 302 vs. 177 days; P , 0.01) that was not ob- SRL172 served in patients with squamous cell cancer (n 5 61) (49, 50). The SRL172 approach relies on the immunogenicity of a heat- killed Mycobacterium vaccae (SRL172) to induce a favorable, GVAX, Allo-GVAX, and Lucanix albeit nonspeciﬁc, systemic immune response that promotes GVAX has been highly publicized through promotion by in- autologous antigen recognition. SR Pharma plc (UK) sponsored dustry sponsors, and the high-proﬁle nature alone makes GVAX a randomized trial using SRL172 concurrently with platinum- and two related studies worthy of discussion. Although none of based chemotherapy with the intention of inducing systemic the agents have undergone randomized, controlled investigation, reactivity during endogenous tumor antigen release. Although the considerable logic of phased development behind this series designed as a phase III, non–placebo controlled, randomized of gene therapy approaches is highly instructive. study, the trial was subject to a high dropout rate, and the GVAX. The initial study evaluated an autologous tumor cell reduced number of evaluable patients limited the statistical vaccine (GVAX) in 33 individuals with advanced- and 10 with power. A total of 419 patients with stage III/IV NSCLC were early-stage NSCLC. Autologous tumor obtained from surgery randomized 1:1 to receive serial injections of 109 bacilli (ﬁve was transduced ex vivo with an adenoviral vector delivering monthly injections followed by monthly maintenance) admin- GM-CSF cDNA to processed tumor cells. Of 33 patients with istered concurrently with six cycles of chemotherapy delivered advanced NSCLC, three (two of which were bronchoalveolar cell on a 21-day schedule. Fewer than 50% of the subjects com- carcinoma) achieved complete response and prolonged remission. pleted the prescribed series of vaccines. Including all subjects, Eight of the 10 patients with early-stage lung cancer remained free the authors concluded that SLR172 signiﬁcantly improved of disease with a median follow-up time of 12 months. The authors patient quality of life without affecting overall survival (median, reported a positive correlation between GM-CSF production by 230 PROCEEDINGS OF THE AMERICAN THORACIC SOCIETY VOL 6 2009 the vaccine and clinical response. An important aspect of this vaccine vs. BSC). A total of 182 patients with resected stage IB/ study was that, of 83 tumors harvested, vaccines could not be II, MAGE-A3–positive NSCLC were serially immunized and successfully generated in 16 patients, and 11 other died before evaluated for time to recurrence. Investigators observed vaccine could be delivered. The median production time was a delayed time to recurrence at 28 months (31.6 [vaccine] vs. 31 days, and median time from tumor harvest to immunization was 43.3% [control]), which translated into a 27% reduction in 49 days. A total of 43 patients were immunized (51). relative risk of cancer recurrence (55, 56). Although differences Allo-GVAX (bystander GVAX). Based on encouraging failed to meet statistical signiﬁcance, results were suggestive anecdotal response that correlated with higher levels of GM- enough to pursue phase III investigation. The international, CSF production, and in response to the limitations imposed by multicenter MAGRIT study (MAGE-A3 Adjuvant Non–Small protracted vaccine production, the group modiﬁed their ap- Cell Lung Cancer Immunotherapy), opened to enrollment in proach. An updated vaccine mixed autologous tumor with an 2007, plans to accrue 2,270 MAGE-A3–positive patients with allogeneic, non–lung cancer cell line (K562 erythroleukemia completely resected stage IB, II, or IIIA NSCLC. cells) engineered to express GM-CSF, and thereby removed the requirement for gene modiﬁcation of individual tumors. The TG4010 vaccine was tested in a nonrandomized study of 86 patients with The TG4010 vaccine uses a dual-gene–based approach to advanced-stage NSCLC, with the objectives of safety, feasibil- coexpress MUC1 antigen and IL-2 (MVA–MUC1–IL-2) with ity, pharmacokinetics, and efﬁcacy. Cell processing was success- a recombinant vaccinia virus. Two serial studies, sponsored by ful in 76 patients, and 49 proceeded to vaccination. Although Transgene (Strasbourg, France) evaluated clinical endpoints in the GM-CSF secretion was 25-fold higher than autologous patients with stage IIIB/IV NSCLC administered TG4010 GVAX, the investigators disappointingly did not observe an concurrently with conventional chemotherapy. The initial phase objective tumor response similar to that seen in the GVAX IIA, nonrandomized study evaluated 44 patients, and observed study. A higher frequency of injection site reactions than with tumor response rate of 37%; 13 partial responses (RECIST), the parent GVAX was also observed (52). 71% with partial response (PR) or SD for over 12 weeks; Lucanix. The same group then evaluated an entirely alloge- estimated median TTP was 6.4 months; estimated overall neic antigen approach that standardized vaccine production and survival was 13 months. In the follow-up phase IIB study, 148 eliminated the need for tumor cell harvest. Lucanix is comprised patients were randomized to TG4010 plus chemotherapy or of four allogeneic NSCLC cell lines transduced with transform- chemotherapy alone (gem/cis). Patients in the vaccine arm were ing growth factor-b antisense plasmid. The study was conducted immunized weekly with TG4010 for 6 weeks, then received as a phase II nonrandomized trial with dose escalation in 75 maintenance injections every 3 weeks until progression. Interim patients with NSCLC (14 early stage, 61 late stage). Endpoints results reported at the American Society of Clinical Oncology included safety, feasibility, pharmacokinetics, and efﬁcacy. in June 2008 indicate statistical differences in clinical response Results reported in 2006 observed 15% response by RECIST, criteria, and a non–statistically signiﬁcant increase in progres- with an increase in estimated survival compared with historical sion-free survival at 6 months. Data on overall survival are not control subjects, and a correlation of survival with higher dose yet mature, but show an encouraging trend favoring survival in levels. In 2008, the group initiated a phase III placebo-controlled the treatment group compared with matched control subjects; trial of Lucanix in stage III and IV NSCLC (53). ﬁnal study results are expected in late 2008 (57). BLP25 (Stimuvax) Phase II Studies: What Have We Learned? BLP25 liposome vaccine (L-BLP25) carries the mucin-1 Supplementing information from early-phase trials, these stud- (MUC1) protein admixed with monophosphoryl lipid A as an ies indicate that immunotherapy carries limited toxicity. Con- immune adjuvant. The randomized, phase IIB trial (1:1 BLP25 sistent with phase II objectives and study design, there is vs. BSC), sponsored by Merck KGaA, enrolled 171 patients suggested beneﬁt in early- and late-stage NSCLC. Data indicate with stage IIIB/IV NSCLC with response or stable disease (SD) that phase III investigation is warranted and necessary. In after ﬁrst-line therapy. Patients, stratiﬁed by stage (IIIB vs. addition to the need for deﬁnitively addressing therapeutic ‘‘wet’’ IIIB/IV), received eight weekly subcutaneous injections efﬁcacy, a number of relevant questions are aﬁeld. These of BLP25 SQ; patients also received additional treatment with include relative potency and comparative efﬁcacy of different 300 mg/m2 cyclophosphamide 3 days before immunization. The formulations, and practical information about necessary dura- primary endpoint was clinical outcomes. Data analysis showed tion of response, ideal dosing schedule, and appropriate timing no statistical difference in overall survival (17.4 vs. 13 mo; P 5 relative to administration of conventional therapies. 0.1), but a strong trend in median survival (30.6 vs; 13.3 mo) in a subgroup of patients with stage IIIB locoregional disease (n 5 FUTURE DIRECTIONS AND OBJECTIVES 35) compared with matched control subjects (n 5 30). Based on the promising results in this subgroup of patients, Merck KGaA The lack of proven clinical beneﬁt continues to encumber sponsored the multicenter (international) phase III, random- development of immunotherapy for lung cancer. This fact ized, double-blind, placebo-controlled trial. The START study makes proven clinical efﬁcacy in phase III trials an overarching (Stimulating Targeted Antigenic Responses to NSCLC) opened objective. Assuming one or more agents are shown to have to enrollment in 2007. Targeted accrual is 1,300 patients with clinical beneﬁt in lung cancer, however small, attention can unresectable stage III NSCLC that have responded to ﬁrst-line, be judiciously turned to optimizing efﬁcacy. There are several platinum-based chemoradiotherapy (54). key points to consider in that pursuit. It is probable that measurable immunological response to relevant tumor antigens MAGE-A3 Antigen-Speciﬁc Cancer Immunotherapy will positively correlate with clinical beneﬁt. In the context of GlaxoSmithKline produced a recombinant MAGE-A3 fusion measurable activity already shown by multiple agents and protein (His-tagged/full-length MAGE-A3 protein/inﬂuenza formulations, it is to be expected that a number of different protein D) plus immune adjuvant AS02B (monophosphoryl agents can yield similar results. Well standardized and validated lipid A and QS21) that was tested in a double-blind, random- immunological assays used as a comparative measure of po- ized, placebo-controlled, phase IIB trial (2:1 randomization, tency may show some differences in biological activity, but it is Hirschowitz and Yannelli: Immunotherapy for Lung Cancer 231 unlikely that there will emerge a single ideal approach. In a search 19. Sandler AB, Dubinett SM. COX-2 inhibition and lung cancer. Semin for the highest-potency vaccine formulations, pragmatic consid- Oncol 2004;3:45–52. erations of simplicity, cost, and portability are highly relevant. 20. Bartlett JB, Dredge K, Dalgleish AG. The evolution of thalidomide and its IMiD derivatives as anticancer agents. Nat Rev Cancer 2004;4:314– Efﬁcacy will, however, still depend on relevance of antigens. 322. Questions remain about whether there are signiﬁcant differences 21. Krieg AM. Toll-like receptor 9 (TLR9) agonists in the treatment of in immunogenicity between autologous and allogeneic tumor cancer. Oncogene 2008;27:161–167. antigen. Similarly, future investigation should address the ther- 22. Emens LA, Jaffee EM. Leveraging the activity of tumor vaccines with apeutic equivalence of various allogeneic approaches, with cytotoxic chemotherapy. Cancer Res 2005;65:8059–8064. practical consideration to the fact that customized monovalent 23. Nowak AK, Lake RA, Robinson BW. Combined chemoimmunotherapy approaches are currently applicable and restricted to a relatively of solid tumours: improving vaccines? Adv Drug Deliv Rev 2006;58: 975–990. small and selected portion of the lung cancer population. It is also 24. Sinkovics JG, Horvath JC. Evidence accumulating in support of cancer becoming clear that abrogation of immune regulatory elements vaccines combined with chemotherapy: a pragmatic review of past will be necessary before immunotherapy achieves full therapeu- and present efforts. Int J Oncol 2006;29:765–777. tic potential (58). Our current knowledge of immune regulation 25. Gulley JL, Madan RA, Arlen PM. Enhancing efﬁcacy of therapeutic in lung cancer already offers multiple options for immune vaccinations by combination with other modalities. Vaccine 2007;25: modulation, and continued investigation promises to yield addi- B89–B96. 26. Zitvogel L, Apetoh L, Ghiringhelli F, Kroemer G. Immunological tional targets and corresponding agents that could improve aspects of cancer chemotherapy. Nat Rev Immunol 2008;8:59–73. efﬁcacy. In the future, it is likely that combinations of multiple 27. Tesniere A, Panaretakis T, Kepp O, Apetoh L, Ghiringhelli F, Zitvogel immunologically active agents, conventional treatment modali- L, Kroemer G. Molecular characteristics of immunogenic cancer cell ties, and novel targeted therapies, used in concert, will overcome death. Cell Death Differ 2008;15:3–12. limitations of any single approach and lead to signiﬁcant 28. Demaria S, Volm MD, Shapiro RL, Yee HT, Oratz R, Formenti SC, improvements in therapeutic outcomes of lung cancer. Muggia F, Symmans WF. Development of tumor-inﬁltrating lympho- cytes in breast cancer after neoadjuvant paclitaxel chemotherapy. Conﬂict of Interest Statement: Neither of the authors has a ﬁnancial relationship Clin Cancer Res 2001;7:3025–3030. with a commercial entity that has an interest in the subject of this manuscript. 29. Iwase M, Watanabe H, Kondo G, Ohashi M, Nagumo M. Enhanced susceptibility of oral squamous cell carcinoma cell lines to FAS- mediated apoptosis by cisplatin and 5-ﬂuorouracil. Int J Cancer 2003; 106:619–625. References 30. Frost P, Ng CP, Belldegrun A, Bonavida B. Immunosensitization of prostate carcinoma cell lines for lymphocyte (CTL, TIL, LAK)- 1. Isobe T, Herbst RS, Onn A. Current management of advanced non– mediated apoptosis via the fas-fas-ligand pathway of cytotoxicity. small cell lung cancer: targeted therapy. Semin Oncol 2005;32:315– Cell Immunol 1997;180:70–83. 328. 31. Yang S, Haluska FG. Treatment of melanoma with 5-ﬂuorouracil or 2. Giaccone G. Epidermal growth factor receptor inhibitors in the treat- ment of non–small-cell lung cancer. J Clin Oncol 2005;23:3235–3242. dacarbazine in vitro sensitizes cells to antigen-speciﬁc CTL lysis 3. Baselga J, Arteaga CL. Critical update and emerging trends in epider- through perforin/granzyme- and Fas-mediated pathways. J Immunol mal growth factor receptor targeting in cancer. J Clin Oncol 2005;23: 2004;172:4599–4608. 2445–2459. 32. Longley DB, Allen WL, McDermott U, Wilson TR, Latif T, Boyer J, 4. Sandler AB, Johnson DH, Herbst RS. Anti-vascular endothelial growth Lynch M, Johnston PG. The roles of thymidylate synthase and p53 in factor monoclonals in non–small cell lung cancer. Clin Cancer Res regulating Fas-mediated apoptosis in response to antimetabolites. 2004;10:4258s–4262s. Clin Cancer Res 2004;10:3562–3571. 5. Dudley ME, Rosenberg SA. Adoptive cell transfer therapy. Semin 33. Yamamoto T, Nagano H, Sakon M, Wada H, Eguchi H, Kondo M, Oncol 2007;34:524–531. Damdinsuren B, Ota H, Nakamura M, Wada H, et al. Partial 6. Raez LE, Fein S, Podack ER. Lung cancer immunotherapy. Clin Med contribution of tumor necrosis factor–related apoptosis-inducing Res 2005;3:221–228. ligand (TRAIL)/TRAIL receptor pathway to antitumor effects of 7. Abu-Shakra M, Buskila D, Ehrenfeld M, Conrad K, Shoenfeld Y. interferon-a/5-ﬂuorouracil against hepatocellular carcinoma. Clin Cancer and autoimmunity: autoimmune and rheumatic features in Cancer Res 2004;10:7884–7895. patients with malignancies. Ann Rheum Dis 2001;60:433–441. 34. Beyer M, Kochanek M, Darabi K, Popov A, Jensen M, Endl E, Knolle 8. Novellino L, Castelli C, Parmiani G. A listing of human tumor antigens PA, Thomas RK, von Bergwelt-Baildon M, Debey S, et al. Reduced recognized by T cells: March 2004 update. Cancer Immunol Immun- frequencies and suppressive function of CD41CD25hi regulatory T other 2005;54:187–207. cells in patients with chronic lymphocytic leukemia after therapy with 9. Pardoll D. Does the immune system see tumors as foreign or self? Annu ﬂudarabine. Blood 2005;106:2018–2025. Rev Immunol 2003;21:807–839. 35. Cross D, Burmester JK. Gene therapy for cancer treatment: past, 10. Hirschowitz EA, Hiestand DM, Yannelli JR. Vaccines for lung cancer. J present and future. Clin Med Res 2006;4:218–227. Thorac Oncol 2006;1:93–104. 36. Keilholz U, Weber J, Finke JH, Gabrilovich DI, Kast WM, Disis ML, 11. Ribas A, Butterﬁeld LH, Glaspy JA, Economou JS. Current develop- Kirkwood JM, Scheibenbogen C, Schlom J, Maino VC, et al. Immu- ments in cancer vaccines and cellular immunotherapy. J Clin Oncol nologic monitoring of cancer vaccine therapy: results of a workshop 2003;21:2415–2432. sponsored by the Society for Biological Therapy. J Immunother 2002; 12. Zou W. Immunosuppressive networks in the tumour environment and 25:97–138. their therapeutic relevance. Nat Rev Cancer 2005;5:263–274. 37. Nagorsen D, Scheibenbogen C, Thiel E, Keilholz U. Immunological 13. Beyer M, Schultze JL. Immunoregulatory T cells: role and potential as monitoring of cancer vaccine therapy. Expert Opin Biol Ther 2004;4: a target in malignancy. Curr Oncol Rep 2008;10:130–136. 1677–1684. 14. Bhattacharya-Chatterjee M, Chatterjee SK, Foon KA. The anti-idiotype 38. Woo EY, Chu CS, Goletz TJ, Schlienger K, Yeh H, Coukos G, Rubin vaccines for immunotherapy. Curr Opin Mol Ther 2001;3:63–69. SC, Kaiser LR, June CH. Regulatory CD4(1)CD25(1) T cells in 15. Dredge K, Marriott JB, Todryk SM, Dalgleish AG. Adjuvants and the tumors from patients with early-stage non–small cell lung cancer and promotion of Th1-type cytokines in tumour immunotherapy. Cancer late-stage ovarian cancer. Cancer Res 2001;61:4766–4772. Immunol Immunother 2002;51:521–531. 39. Wolf AM, Wolf D, Steurer M, Gastl G, Gunsilius E, Grubeck- 16. Pashine A, Valiante NM, Ulmer JB. Targeting the innate immune Loebenstein B. Increase of regulatory T cells in the peripheral blood response with improved vaccine adjuvants. Nat Med 2005;11:S63–S68. of cancer patients. Clin Cancer Res 2003;9:606–612. 17. Smyth MJ, Cretney E, Kershaw MH, Hayakawa Y. Cytokines in cancer 40. Okita R, Saeki T, Takashima S, Yamaguchi Y, Toge T. CD41CD251 immunity and immunotherapy. Immunol Rev 2004;202:275–293. regulatory T cells in the peripheral blood of patients with breast 18. Cranmer LD, Trevor KT, Hersh EM. Clinical applications of dendritic cancer and non–small cell lung cancer. Oncol Rep 2005;14:1269–1273. cell vaccination in the treatment of cancer. Cancer Immunol Immun- 41. Audia S, Nicolas A, Cathelin D, Larmonier N, Ferrand C, Foucher P, other 2004;53:275–306. Fanton A, Bergoin E, Maynadie M, Arnould L, et al. Increase of 232 PROCEEDINGS OF THE AMERICAN THORACIC SOCIETY VOL 6 2009 CD41 CD251 regulatory T cells in the peripheral blood of patients 50. Stanford JL, Stanford CA, O’Brien ME, Grange JM. Successful immu- with metastatic carcinoma: a phase I clinical trial using cyclophos- notherapy with Mycobacterium vaccae in the treatment of adenocar- phamide and immunotherapy to eliminate CD41 CD251 T lympho- cinoma of the lung. Eur J Cancer 2008;44:224–227. cytes. Clin Exp Immunol 2007;150:523–530. 51. Nemunaitis J, Sterman D, Jablons D, Smith JW II, Fox B, Maples P, 42. Sakaguchi S, Sakaguchi N, Shimizu J, Yamazaki S, Sakihama T, Itoh M, Hamilton S, Borellini F, Lin A, Morali S, et al. Granulocyte- Kuniyasu Y, Nomura T, Toda M, Takahashi T. Immunologic toler- macrophage colony–stimulating factor gene-modiﬁed autologous tumor ance maintained by CD251 CD41 regulatory T cells: their common vaccines in non–small-cell lung cancer. J Natl Cancer Inst 2004;96: role in controlling autoimmunity, tumor immunity, and transplanta- 326–331. tion tolerance. Immunol Rev 2001;182:18–32. 52. Nemunaitis J, Jahan T, Ross H, Sterman D, Richards D, Fox B, Jablons 43. Sakaguchi S. Naturally arising CD41 regulatory t cells for immunologic D, Aimi J, Lin A, Hege K. Phase 1/2 trial of autologous tumor mixed self-tolerance and negative control of immune responses. Annu Rev with an allogeneic GVAX vaccine in advanced-stage non–small-cell Immunol 2004;22:531–562. lung cancer. Cancer Gene Ther 2006;13:555–562. 44. Jaffe CC. Measures of response: RECIST, WHO, and new alternatives. 53. Nemunaitis J, Dillman RO, Schwarzenberger PO, Senzer N, Cunning- J Clin Oncol 2006;24:3245–3251. ham C, Cutler J, Tong A, Kumar P, Pappen B, Hamilton C, et al. 45. Giaccone G, Debruyne C. Phase III study of BEC/BCG vaccination in Phase II study of belagenpumatucel-L, a transforming growth factor limited disease small cell lung cancer patients following response to b-2 antisense gene-modiﬁed allogeneic tumor cell vaccine in non– chemotherapy and thoracic irradiation (EORTC 08971, the SILVA small-cell lung cancer. J Clin Oncol 2006;24:4721–4730. study). J Clin Oncol 2004;22:622s. 54. Butts C, Murray N, Maksymiuk A, Goss G, Marshall E, Soulieres D, ` 46. Giaccone G, Debruyne C, Felip E, Chapman PB, Grant SC, Millward M, Cormier Y, Ellis P, Price A, Sawhney R, et al. Randomized phase IIB Thiberville L. D’addario G, Coens C, Rome LS, et al.. Phase III study trial of BLP25 liposome vaccine in stage IIIB and IV non–small-cell ´rin of adjuvant vaccination with BEC2/bacille Calmette-Gue in respond- lung cancer. J Clin Oncol 2005;23:6674–6681. ing patients with limited-disease small-cell lung cancer (European 55. Vansteenkiste J, Zielinski M, Linder A, Dahabre J, Esteban E, Organization for Research and Treatment of Cancer 08971–08971B; Malinowski W, Jassem J, Passlick B, Lehmann F, Brichard VG. Silva Study). J Clin Oncol 2005;23:6854–6864. Final results of a multi-center, double-blind, randomized, placebo- 47. ´ ´ Garcıa B, Neninger E, de la Torre A, Leonard I, Martınez R, Viada C, controlled phase II study to assess the efﬁcacy of MAGE-A3 im- ´lez Gonza G, Mazorra Z, Lage A, Crombet T. Effective inhibition of the munotherapeutic as adjuvant therapy in stage IB/II non-small cell epidermal growth factor/epidermal growth factor receptor binding by lung cancer (NSCLC) [abstract]. J Clin Oncol 2007;25:18S:7554. anti-epidermal growth factor antibodies is related to better survival in 56. Brichard VG, Lejeune D. GSK’s antigen-speciﬁc cancer immunotherapy advanced non–small-cell lung cancer patients treated with the epider- programme: pilot results leading to phase III clinical development. mal growth factor cancer vaccine. Clin Cancer Res 2008;14:840–846. Vaccine 2007;25(suppl 2):B61–B71. 48. ´ Neninger Vinageras E, de la Torre A, Osorio Rodrıguez M, Catala ´ 57. Ramlau R, Westeel V, Papai Z, Riviere A, Madroszyk A, Koralewski P, Ferrer M, Bravo I, Mendoza del Pino M, Abreu Abreu D, Acosta Lacoste G, Acres B, Limacher JM, Quoix E. Randomized phase IIb trial Brooks S, Rives R, del Castillo Carrillo C, et al. Phase II randomized evaluating the therapeutic vaccine TG4010 (MVA-MUC1-IL2) as an controlled trial of an epidermal growth factor vaccine in advanced adjunct to chemotherapy in patients with advanced non–small cell lung non–small-cell lung cancer. J Clin Oncol 2008;26:1452–1458. cancer (NSCLC) [abstract]. J Clin Oncol 2008 [accessed 2008 Oct 2]. Avail- 49. O’Brien ME, Anderson H, Kaukel E, O’Byrne K, Pawlicki M, Von able from: http://www.asco.org/ASCO/Abstracts1&1Virtual1Meeting/ Pawel J, Reck M. SR-ON-12 Study Group. SRL172 (killed Mycobac- Abstracts?&vmview5abst_detail_view&confID555&abstractID532916. terium vaccae) in addition to standard chemotherapy improves quality 58. Laheru DA, Pardoll DM, Jaffee EM. Genes to vaccines for immuno- of life without affecting survival, in patients with advanced non–small- therapy: how the molecular biology revolution has inﬂuenced cancer cell lung cancer: phase III results. Ann Oncol 2004;15:906–914. immunology. Mol Cancer Ther 2005;4:1645–1652.