Stimulation of a host immune response
Nonspecific Immune Stimulation
Previous attempts at nonspecific stimulation of the immune response, such as direct intratumor injections of bacillus Cal-mette-Guerin (BCG) or other immune adjuvant compounds, often provoked local immune responses (with occasional regressions of injected tumors) but without consistent systemic effects on noninjected sites of disease. The local administration of BCG has been successful in treating and preventing the recurrence of noninvasive bladder cancer [see 12:XIV Bladder, Renal, and Testicular Cancer].7
Vaccine Therapy
A major strategy for generating a systemic immune response against cancer is the development of tumor vaccine therapy.8 The goal of such therapy is to present tumor-associated antigens in association with appropriate costimulatory molecules in a way that circumvents the tumor-evasive mechanisms that would ordinarily lead to tolerance (a state of immune paralysis). Approaches to immunizing a tumor-bearing host with tumor-associated antigens have ranged from vaccination with intact tumor cells to administration of selected immunogenic peptide vaccines.
Vaccination with intact unmodified tumor cells Even with adjuvant compounds, vaccination with intact unmodified tumor cells has generally been ineffective in triggering a clinically relevant immune response. If tumor cells escape immune recognition in situ, immunization with cells from the tumor carries the same limitations.
Vaccination with genetically modified tumor cells In this strategy, host-derived tumor cells are transfected with genes encoding for allogeneic MHC genes, costimulatory molecules (e.g., B7), or immunomodulatory cytokines. These genetically modified tumor cells bearing MHC or costimulatory molecules compensate for the lack of these molecules in vivo. Tumor cells trans-fected with genes encoding for immunomodulatory cytokines may result in the release of these soluble factors (e.g., granulo-cyte-macrophage colony-stimulating factor [GM-CSF], IL-4, and IL-2), which recruit dendritic cell precursors or helper T cells into vaccination sites to facilitate the immune recognition of tumor antigen and the subsequent generation of an effector cytotoxic T lymphocyte (CTL) response. Rather than relying on the cumbersome technique of transfecting autologous tumor cells in vitro, a more practical approach is to admix autologous tumor cells with generic transduced bystander cells or biopolymer microspheres containing cytokines or to use in situ intratumoral injection of genetic material. An alternative approach is to employ standardized (allogeneic) gene-transduced tumor cell lines that display broadly expressed antigens as a source of tumor vaccine.
Vaccination with tumor antigens or peptides The goal of this strategy is to immunize patients with tumor antigens or immunodominant peptides derived from those antigens. An example is the direct immunization with a peptide derived from the MART-1 antigen as a means of provoking an immune response from patients with melanoma. Several class I (HLA-A2)-restricted, highly immunogenic peptides of MART-1 and GP-100 have been created.9,10 In one clinical trial, 42 patients with metastatic melanoma were vaccinated with a synthetic peptide derived from GP-100.11 Eleven patients receiving the GP-100-derived peptide without IL-2 demonstrated no objective responses, but 13 of 31 patients (42%) receiving the peptide plus systemic IL-2 achieved a clinically relevant tumor shrinkage. The results of this early investigation are consistent with the concept that vaccines for treatment of cancer will need to be used along with adjuvants to enhance immunologic response (i.e., IL-2 or other signaling molecules).
A similar vaccination strategy was evaluated in 10 patients with multiple myeloma.12 The vaccine consisted of autologous immunoglobulin idiotype (Id) protein purified from serum or urine, coupled to keyhole limpet hemocyanin (KLH), a protein commonly utilized as an adjuvant to augment an immune response. The patients received subcutaneous injections of Id-KLH conjugates along with low doses of GM-CSF or IL-2. Although the clinical impact of this immunization remains to be determined, 8 of 10 patients developed an Id-specific delayed-type hypersensitivity (DTH) reaction, indicating the immunocompe-tence of multiple myeloma patients for specific immunization.
Vaccination with recombinant viral or bacterial vaccines
Another strategy for stimulating an immune response to a specific tumor antigen is through the use of recombinant viral or bacterial vaccines. In this approach, the viral or bacterial genome is modified to include a gene encoding the relevant tumor antigen with or without genes encoding for costimulatory molecules (e.g., B7). The recombinant viral or bacterial agents then act as vectors with the capacity of infecting APCs or tumor cells. This strategy also takes advantage of the natural adjuvant effect and intrinsic immunogenicity of viruses and bacteria.
Naked DNA vaccines Yet another strategy involves vaccination with so-called naked DNA encoding for a tumor-associated antigen. These nucleic acid vaccines may be less potent than vaccines comprising recombinant virus or bacteria, because unlike the latter, they have no mechanism for amplification in the host individual.
Dendritic cell vaccines The use of DCs as a major constituent of a tumor vaccine helps ensure effective presentation of antigen to CD4+ and CD8+ T cells.13 Various approaches have been devised to load DCs with tumor-associated antigen, including (1) ex vivo incubation of autologous DCs with tumor antigen or tumor cell lysates before vaccination, (2) fusion of DCs with whole autologous tumor cells, and (3) ex vivo trans-duction with RNA or replication-defective recombinant viral vectors carrying tumor-associated antigens within their genome. An example of DC-based vaccination strategy has been evaluated in B cell non-Hodgkin lymphoma.14 Autolo-gous DCs were cocultured with idiotypic protein from the patient’s malignant B cells and then utilized as a vaccine. Of the 35 patients vaccinated, the majority developed T cell or humoral anti-Id responses. Among the 28 patients with measurable disease at the time of vaccination (or subsequent booster injections), 11 experienced clinically significant tumor regressions, including four with complete responses (i.e., no clinically detectable tumor).
Systemic Administration of Immunostimulatory Cytokines
Another approach to stimulating an immune response in a patient with cancer is to administer recombinant cytokines that may enhance an otherwise ineffective immune response. An example of this strategy is the use of high-dose IL-2, which enhances the immune response by stimulating T cell and NK cell activity and increases the serum concentration of other im-munostimulatory cytokines, such as TNF, IL-1, and interferon gamma. Systemic high-dose IL-2 has been extensively evaluated in the treatment of metastatic melanoma and renal cell carcinoma and is approved by the Food and Drug Administration for the treatment of these two diseases. The approval of high-dose IL-2 as therapy for metastatic melanoma is based in part on the collective experience of 270 patients who were entered into eight clinical trials in the United States. Doses of 600,000 to 720,000 IU/kg I.V. were administered every 8 hours over two 5-day periods, with a 6- to 9-day rest in between the 5-day cycles. Partial tumor shrinkage occurred in 10% of the treated patients, with another 6% demonstrating complete tumor shrink-age.15 For the patients who experienced complete response, the median response duration was not reached, and 59% of these patients remained progression free. Similar response rates and durations of response were seen in patients with metastatic renal cell carcinoma.16 The major drawback to this form of therapy is its associated toxicity, which includes the development of a capillary leak/shock syndrome.
Interferon alfa, which has direct antiproliferative effects and causes increased MHC class I expression, has been effective in inducing tumor regressions in a minority of patients with renal cell carcinoma, melanoma, hairy-cell leukemia, and lym-phoma; it has also been effective in the treatment of chronic myelogenous leukemia. Other immunomodulatory cytokines that have been evaluated in clinical trials include TNF-a and interferon gamma, both of which have shown antitumor activity in vitro but are poorly tolerated or minimally active in vivo.
Passive immunity against cancer by adoptive transfer of cells or antibody
The strategy of adoptive transfer of preimmune cells or antibody to a tumor-bearing patient circumvents the potential limitation of stimulating an effective immune response in favor of boosting the host with previously activated elements of an immune effector response.
Adoptive Cellular Immunotherapy
Adoptive cellular immunotherapy is the transfer of ex vivo activated immune cells with antitumor activity into a tumor-bearing host. Examples of this strategy include the adoptive transfer of lymphokine-activated killer (LAK) cells. In this approach, peripheral blood lymphoid cells (primarily NK cells) are taken from patients with cancer and are subsequently activated and expanded in vitro with IL-2; they are then administered back to the patient (usually along with systemic IL-2). Initial results of the use of LAK cells with high-dose IL-2 in patients with renal cell carcinoma and melanoma were encouraging, but subsequent randomized trials failed to show any benefit of LAK cell administration over the modest response rate seen with high-dose IL-2 alone.17 As an alternative to peripheral blood lymphocytes, investigators have evaluated the adoptive transfer of tumor-infiltrating lymphoid (TIL) cells isolated from inflammatory infiltrates of excised tumor nodules. Once expanded in IL-2, these cells may be enriched for tumor-specific cytotoxic T cells and NK cells. Of 86 patients with stage IV melanoma receiving TIL treatment at the National Cancer Institute, 34% had objective tumor responses (i.e., partial or complete tumor regressions).18 Unfortunately, it was not possible to isolate sufficient numbers of TIL cells from all tumor specimens to accomplish adoptive transfer. Other patients experienced disease progression while their TIL cells were being cultured and therefore never received treatment. Despite the modest success of TIL therapy for melanoma, comparable results have not been observed in the treatment of patients with renal cell carcinoma. One of the side benefits of TIL technology has been its use in isolating and defining the tumor antigens recognized by T cells from TIL cell cultures, such as MART-1, GP-100, tyrosinase, p15, TRP-1, and |-catenin, which are all expressed by melanoma cells.19
Other sources for adoptively transferred immune effector cells include the ex vivo activation and expansion of lymphoid cells derived from draining lymph nodes at tumor vaccination sites or in situ genetic modification of tumor nodules (direct injection of genetic material that encodes for proteins into tumors, triggering an immune response).
Genetic-modification strategies have been used to enhance the cytolytic activity of adoptively transferred immune effector cells. This technique involves the transduction of autologous CTLs (or their bone marrow precursors) with genes encoding a TCR specific for a tumor antigen (e.g., a TCR that recognizes MART-1 antigen in melanoma) or involves a chimeric receptor (i.e., an antigen-binding domain linked to a signal-transducing domain that initiates cellular activation after cross-linking by antigen) recognizing a tumor-associated antigen such as the mucin antigen TAG-72 expressed by most adenocarcinomas.20,21 Mononuclear phagocytes have also been evaluated as a source of immune effector cells for adoptive cellular immuno-therapy. This strategy has involved the ex vivo activation of peripheral blood monocyte-derived macrophages by interferon gamma or lipopolysaccharide, with the subsequent administration of these activated cells back into the tumor-bearing host. However, as with other adoptive cellular therapy strategies, there has been minimal therapeutic response to date.22
Another form of adoptive cellular therapy that has received considerable attention is the use of allogeneic donor lymphocyte infusions. The rationale for this strategy is the fact that in allo-geneic bone marrow/stem cell transplantation, donor lymphocytes appear to exert a graft-versus-leukemia effect. This approach is supported by data showing a higher risk of leukemic relapse in patients receiving syngeneic bone marrow transplantation than in patients receiving allogeneic bone marrow transplantation, as well as a higher risk of relapse in patients receiving T cell-depleted bone marrow than in those receiving unmanipulat-ed allogeneic bone marrow.23 The greatest benefit of allogeneic lymphocyte infusions was seen in patients with chronic myeloid leukemia (CML) who had relapses after allogeneic bone marrow transplantation; of these patients, 70% had complete responses to donor lymphocyte infusions.24 This donor-lymphocyte graft-ver-sus-leukemia effect may reflect immune reactivity toward minor histocompatibility antigens shared between recipient tissues (resulting in graft-versus-host disease) and leukemic cells, as well as reactivity toward leukemia-associated antigens. The efficacy of donor-lymphocyte infusions seen in patients with CML has led to the use of such infusions in other diseases, and investigators are currently assessing the value of the graft-versus-tumor effect as a primary form of antineoplastic therapy. In this strategy, a nonmyeloablative (i.e., not causing sustained neutropenia) im-munosuppressive preparative regimen is used to achieve en-graftment of allogeneic peripheral blood or bone marrow-derived stem cells, with the express purpose of developing a graft-versus-tumor effect. After allogeneic stem cell transplant engraftment, additional infusions of donor lymphocytes are often used to enhance the graft-versus-tumor effect. When this approach was expanded to include patients with chronic lympho-cytic leukemia (CLL) or indolent non-Hodgkin lymphoma (NHL), promising results were observed.25 In one study involving 15 patients with CLL or indolent NHL, 11 patients were engrafted, and all 11 had objective responses; eight patients had complete responses. The encouraging results seen in patients with hematologic malignancies have precipitated clinical trials to determine the feasibility and efficacy of this approach in patients with solid tumors. Preliminary results indicate that this approach has some efficacy in renal cell carcinoma.
Antitumor Antibody Therapy
The administration of antitumor antibodies to patients with cancer may result in tumor regression by a variety of mechanisms, including ADCC caused by NK cells and macrophages; cytotoxicity resulting from complement fixation; induction of apoptosis; and the delivery of toxins, cytotoxic pharmaceutical agents, or ionizing radiation to the tumor. In the treatment of lymphoma, promising results have been observed with the use of monoclonal antibodies that are specific for surface idiotype and for the CD20 differentiation antigen. An FDA-approved therapy for NHL employs a monoclonal antibody targeting an 85 kd B cell differentiation antigen (B1, CD20), which is expressed by most normal and malignant B cells but not by stem cells or plasma cells. The advantage of CD20 as a target of im-munotherapy is that it is not shed, internalized, or otherwise modulated as a result of antibody binding. Its mechanisms of B cell cytotoxicity may include complement fixation, ADCC, and the transmission of apoptotic signals. Intravenous administration of the chimeric human/mouse CD20 antibody rituximab, alone or in combination with chemotherapy, has produced significant disease regression in patients with low-grade or follic-ular NHL. In a phase 3 trial of 166 patients who received four doses of rituximab alone after having relapses of indolent lym-phoma,27 48% of the treated patients achieved a clinically relevant response, including 6% who experienced complete tumor shrinkage. The median time to progression was 13 months. In another study, in which rituximab was used to treat 28 evalu-able patients with bulky relapsed NHL (the patients had a large tumor burden),28 the response rate was 43%, and the median time to progression was 8.1 months. Rituximab may show its greatest clinical utility when used in conjunction with chemotherapy. Of 35 evaluable patients with NHL who were not previously treated, 100% of the patients experienced a response to combination therapy; 63% experienced complete re-sponse.29 Seven of eight evaluable patients had a so-called molecular complete response, as evidenced by the loss of the bcl-2 oncogene on polymerase chain reaction analysis.
An alternative approach for the use of anti-CD20 antibody in the treatment of patients with low-grade or follicular NHL employs murine antibodies radioconjugated to either yttrium-90 ibritumomab (90Y-ibritumomab) or iodine-131 tositumomab (131I-tositumomab). In two therapeutic trials,30,31 a single intravenous dose of 90Y-ibritumomab given to patients with relapsed or refractory low-grade or follicular NHL produced overall response rates of 74% to 80% (15% to 30% had complete responses); the median time to progression was 6.8 to 11.2 months. Similarly, a single intravenous dose of 131I-tositumomab given to a comparable group of NHL patients caused significant clinical responses in 57% to 65%; 17% to 32% of these patients achieved complete shrinkage of their visible tumor. The median duration of response was 20 months.32,33 In previously untreated patients with follicular lymphoma, treatment with 131I-tositumomab produced an overall response rate of 97%; 63% of these patients achieved a complete response. The rate of progression-free survival was 68% at 3 years.34 These promising findings recently led to FDA approval of 90Y-ibritumomab for the treatment of patients with relapsed or refractory low-grade or follicular NHL; FDA approval of 131I-tositumomab is pending.
Antitumor antibody therapy has also proved effective against breast cancer. The antigenic target of an FDA-approved breast cancer treatment is HER-2/neu, the 185 kd transmembrane growth factor receptor that is overexpressed in 25% of human breast cancers (and in a variety of other malignancies). Preclini-cal studies demonstrated that anti-HER-2/neu monoclonal antibody exposure of cells overexpressing HER-2/neu had antipro-liferative activity and may facilitate apoptotic cell death. On the basis of these promising preclinical findings, trastuzumab (Her-ceptin), a form of anti-HER-2/neu monoclonal antibody, was evaluated alone and in combination with chemotherapy in patients with metastatic breast carcinoma.35,36 In 222 patients receiving weekly intravenous trastuzumab alone, the overall response rate was 14%, with 2% having complete response; the median duration of response was 9.1 months. In a separate clinical trial, 469 patients were randomized to receive either trastuzumab plus chemotherapy (with doxorubicin-cyclophosphamide or pac-litaxel) or chemotherapy alone. The combination of trastuzumab and chemotherapy was superior to chemotherapy alone with regard to time to progression, overall response rate, and 1-year survival. These promising findings led the FDA to approve trastuzumab as an adjunct to chemotherapy in patients with HER-2/neu-positive breast cancers.
Antitumor antibody therapy has been applied to the treatment of patients with colorectal cancer. The antigenic target in this disease is a nonsecreted 40 kd glycoprotein recognized by the investigational 17-1A monoclonal antibody. The 17-1A antigen is overexpressed by most epithelioid tumors, including the majority of colorectal carcinomas. Although the 17-1A antibody displayed only minor to modest clinical activity when administered intravenously to patients with advanced stage IV disease, it may be more active as adjuvant therapy for patients who are at high risk for recurrence after surgical resection. In a study of 189 patients with stage III colorectal carcinoma (i.e., patients found to have local lymph node metastases) who were randomly assigned either to observation alone or to four post operative infusions of 17-1A monoclonal antibody,37 mortality in the treatment arm was reduced by 32%, and recurrence was decreased by 23%.
Other strategies of antitumor-antibody therapy include the use of toxin or drug-conjugated monoclonal antibodies targeted to tumor-associated antigens (in addition to radionuclide-conjugated antibody treatment). Another approach involves the administration of bispecific (heteroconjugate) antibodies consisting of two covalently linked antibodies, with one antigen recognition site targeted to the target tumor-associated antigen and the other recognition site targeted to a trigger molecule (expressed on the effector cell).38 Antitumor antibodies may also be useful as a means of depleting tumor cells contaminating preparations of autologous bone marrow cells before transplantation.39
Immunotherapeutic approaches to the treatment of cancer is an area of active investigation. Most of these approaches are in-vestigational and continue to be evaluated in clinical trials. However, several of the therapeutic agents (IL-2, trastuzumab, BCG, and rituximab) have been shown to be effective and are approved by the FDA.
Diagnostic Role of Monoclonal Antibodies
Diagnosis of leukemias and lymphomas
In addition to their potential role as therapeutic agents, monoclonal antibodies that recognize tumor-associated antigens are being used to facilitate the diagnosis of hematologic malignancies.40,41 The rationale for their use as diagnostic tools stems from the fact that leukemias and lymphomas are thought to be neoplastic counterparts of subpopulations of normal lymphoid and myeloid cells. The identification of cell surface antigens common to both normal and neoplastic lym-phoid and myeloid cells has made it possible to assign the normal cell lineage to the malignant counterparts. This lineage information provides an additional means of discriminating among the various forms of leukemia and lymphoma that may influence treatment decisions. An example of this concept is T cell differentiation [see Figure 3]. Research over the past 20 years has led to the identification of multiple cell surface antigens expressed by cells of the T cell lineage that are differentially expressed along the path of differentiation. Monoclonal antibodies have been developed for these so-called differentiation antigens and have been utilized in the diagnosis of leukemia and lymphoma. CD34 and MHC class II gene products, for example, are expressed by the prethymocyte but are lost by thymic T cell precursors. Conversely, in the thymus, CD4 and CD8 accessory molecules are coexpressed by a biphenotypic T cell precursor that gives rise to separate CD4 and CD8 lineages in the thymic medulla, with subsequent maturation of these lineages in peripheral lymphoid tissues. T cell acute lymphoblastic leukemia/lymphoma shares the surface marker phenotypic characteristics of the bone marrow prethymocyte and is therefore believed to be a malignant transformation of this normal stem cell. The T cell lymphomas and chronic leukemias, on the other hand, share the cell surface phenotype characteristic of mature CD4 or CD8 lymphoid cells found in the peripheral lymphoid tissues. They, in turn, are thought to represent malignant transformation of these more mature T cells.
Figure 3 Sequential expression of selected antigens during T cell development. Depicted is the pathway of T cell differentiation from the bone marrow stem cell to thymocyte maturation in the thymus, leading to the diversion of the CD4+ helper and CD8 cytotoxic suppressor cell sublineages, each of which undergoes further differentiation in peripheral lymphoid tissues. Certain disorders are shown under the associated phenotypes. (DR—HLA class II; T-ALL —T cell acute lymphoblastic leukemia; TdT— terminal deoxynucleotidyl transferase)
The expression of cell surface markers of leukemias and lymphomas can have diagnostic and prognostic significance:
1. Patients with T cell acute lymphoblastic leukemia whose cells express markers characteristic of the earlier stage of thymocyte development have a higher rate of treatment failure during induction than patients whose cells express a more mature phenotype.42
2. One third of patients with acute lymphoblastic leukemia (ALL) have leukemic cells expressing myeloid markers (biphenotypic ALL). Adult patients with biphenotypic ALL have a poor response to therapy and a shorter survival.43
3. Acute leukemias that are not classified by morphology or cytochemical characteristics (approximately 1%) can often be classified as belonging to a lymphoid lineage, as opposed to a nonlymphoid lineage, through the use of cell surface markers. In classifying myeloid leukemias into subgroups, analysis of cell surface markers can supplement morphology and cytochemical staining.
4. Of patients with CML in blast crisis, approximately one third express lymphocyte markers (e.g., CD10) and preferentially respond to the therapy for ALL.
Diagnosis and staging of cancer in vivo
In addition to their potential utility in the pathologic diagnosis of leukemias and lymphomas, monoclonal antibodies have been employed in the diagnosis and staging of solid tumors.44 Monoclonal antibodies specific for tumor-associated antigens have been useful in the diagnosis of carcinomas of unknown origin.
Using a so-called cocktail of antibodies, pathologists are better able to determine the organ of origin of a metastatic carcinoma; such determinations may affect treatment decisions. In the setting of colon cancer, the CYT-103 (OncoScint) antibody has been approved by the FDA for use as a supplemental tool for the staging of colon cancer. The CYT-103 scan incorporates Indium-111-labeled CYT-103 antibody that is specific for a tumor-associated glycoprotein found on many mucin-containing adenocarci-nomas, including many colon cancers. When used as a diagnostic tool, the CYT-103 antibody has a reported sensitivity of 92% and a specificity of 67%; its use results in a change in patient management in 33% of patients evaluated.45 Its use may be superior to CT scanning in the detection of pelvic tumors and ex-trahepatic abdominal metastases. In addition, the CYT-103 antibody and other antibodies specific for the oncofetal protein CEA have been employed to identify the occult sites of metastatic disease in patients in whom the CEA level is rising, in the absence of radiographic evidence of disease recurrence. The CYT-103 antibody has also been used to detect recurrent ovarian carcinoma in the peritoneal cavity. A similarly tagged monoclonal antibody specific for prostate-specific membrane antigen is being utilized in the staging of prostate cancer.46 Radiolabeled monoclonal antibodies specific for other cancers are being developed and evaluated for their utility in the management of cancer.
