Tumor endothelial marker 8 enhances tumor immunity in conjunction with immnization against differentiation Ag.

BACKGROUND: We have previously shown that xenogenic DNA vaccines encoding rat neu and melanosomal differentiation Ag induce tumor immunity. Others have developed vaccines targeting tumor neovasculature. Tumor endothelial marker 8 (TEM8) is expressed in the neovasculature of human tumors, and in the mouse melanoma B16, but its expression is limited in normal adult tissues. We describe a DNA vaccine combining xenogeneic tumor Ag and TEM8. METHODS: In-situ hybridization was used to detect TEM8 RNA in mouse tumors. Mice transgenic for the rat neu proto-oncogene were immunized with DNA vaccines encoding TEM8 and the extracellular domain of rat neu and challenged with the 233-VSGA1 breast cancer cell line. In parallel experiments, C57BL/6 mice were immunized with TEM8 and human tyrosinase-related protein 1 (hTYRP1/hgp75) and challenged with B16F10 melanoma. RESULTS: TEM8 was expressed in the stroma of transplantable mouse breast and melanoma tumors. In both model systems, TEM8 DNA had no activity as a single agent but significantly enhanced the anit-tumor immunity of neu and hTYRP1/hgp75 DNA vaccines when given in concert. The observed synergy was dependent upon CD8+ T cells, as depletion of this cell population just prior to tumor challenge obviated the effect of the TEM8 vaccine in both tumor models. DISCUSSION: A local immune responses to TEM8 may increase inflammation or tumor necrosis within the tumor, resulting in improved Ag presentation of HER2/neu and hTYRP1/hgp75. Alternatively, TEM8 expression in host APC may act more as an adjuvant than an immunologic target.

Tumor endothelial marker 8 enhances tumor immunity in conjunction with immunization against differentiation Ag.

BACKGROUND: We have previously shown that xenogeneic DNA vaccines encoding rat neu and melanosomal differentiation Ag induce tumor immunity. Others have developed vaccines targeting tumor neovasculature. Tumor endothelial marker 8 (TEM8) is expressed in the neovasculature of human tumors, and in the mouse melanoma B16, but its expression is limited in normal adult tissues. We describe a DNA vaccine combining xenogeneic tumor Ag and TEM8. METHODS: In-situ hybridization was used to detect TEM8 RNA in mouse tumors. Mice transgenic for the rat neu proto-oncogene were immunized with DNA vaccines encoding TEM8 and the extracellular domain of rat neu and challenged with the 233-VSGA1 breast cancer cell line. In parallel experiments, C57BL/6 mice were immunized with TEM8 and human tyrosinase-related protein 1 (hTYRP1/hgp75) and challenged with B16F10 melanoma. RESULTS: TEM8 was expressed in the stroma of transplantable mouse breast and melanoma tumors. In both model systems, TEM8 DNA had no activity as a single agent but significantly enhanced the anti-tumor immunity of neu and hTYRP1/hgp75 DNA vaccines when given in concert. The observed synergy was dependent upon CD8+ T cells, as depletion of this cell population just prior to tumor challenge obviated the effect of the TEM8 vaccine in both tumor models. DISCUSSION: A local immune response to TEM8 may increase inflammation or tumor necrosis within the tumor, resulting in improved Ag presentation of HER2/neu and hTYRP1/hgp75. Alternatively, TEM8 expression in host APC may alter T-cell interactions or homing. In this way, TEM8 may act more as an adjuvant than an immunologic target.

In vitro expansion of Ag-specific T cells by HLA-A*0201-transfected K562 cells for immune monitoring.

BACKGROUND: Development of a practical and sensitive assay for evaluating immune responses against cancer Ag has been a challenge for immune monitoring of patients. We have established a reproducible method using peptide-pulsed K562-A*0201 cells as APC to expand Ag-specific T cells in vitro. This method may be applied for monitoring T-cell responses in cancer immunotherapy clinical trials. METHODS: Autologous PBMC from HLA-A*0201+ healthy donors and patients with melanoma were stimulated with peptide-pulsed K562-A*0201 cells under varying conditions. We investigated (1) different culture conditions, including the requirements for serum and cytokines for expansion of CD8+ T lymphocytes; (2) a range of peptide concentrations for Ag loading; (3) phenotypic characterization of responding T cells; and (4) APC:responder ratios and their effects on T-cell expansion. We validated these conditions by ELISPOT and intracellular cytokine staining (ICS) assays using peptides from influenza, Epslein-Barr Virus (EBV) and tyrosinase. RESULTS: Conditions for optimal T-cell expansion using K562-A*0201 APC included input of 2 x 10(6) PBMC, a 10 microg/mL peptide concentration to pulse K562-A*0201 cells, a 1:30 APC:responder T-cell ratio and culture in 10% autologous plasma supplemented with IL-2 and IL-15. In these conditions, Ag-specific T cells expanded >100-fold over a 10-day culture period (peak at day 12). DISCUSSION: This bulk culture method is simple and reliable for expanding human Ag-specific T cells using peptide-pulsed K562-A*0201 cells. This HLA-matched APC line can be adapted to other HLA haplotypes, and has advantages for monitoring clinical trials of immunotherapy with limited availability of autologous APC and PBMC from patients.

Development of a xenogeneic DNA vaccine program for canine malignant melanoma at the Animal Medical Center.

INTRODUCTION: Canine malignant melanoma (CMM) is an aggressive neoplasm treated with surgery and/or fractionated RT; however, metastatic disease is common and chemoresistant. Preclinical and clinical studies by our laboratory and others have shown that xenogeneic DNA vaccination with tyrosinase family members can produce immune responses resulting in tumor rejection or protection and prolongation of survival. These studies provided the impetus for development of a xenogeneic DNA vaccine program in CMM. MATERIALS AND METHODS: Cohorts of three dogs each received increasing doses of xenogeneic plasmid DNA encoding either human tyrosinase (huTyr; 100/500/1500 mcg), murine GP75 (muGP75; 100/500/1500 mcg), murine tyrosinase (muTyr; 5 dogs each at 100/500 mcg), muTyr+/-HuGM-CSF (9 dogs at 50 mcg muTyr, 3 dogs each at 100/400/800 mcg HuGM-CSF, or 3 dogs each at 50 mcg muTyr with 100/400/800 mcg HuGM-CSF), or 50 mcg MuTyr intramuscularly biweekly for a total of four vaccinations. RESULTS: The Kaplan-Meier median survival time (KM MST) for all stage II-IV dogs treated with huTyr, muGP75 and muTyr are 389, 153 and 224 days, respectively. Preliminarily, the KM MST for stage II-IV dogs treated with 50 mcg MuTyr, 100/400/800 mcg HuGM-CSF or combination MuTyr/HuGM-CSF are 242, 148 and >402 (median not reached) days, respectively. Thirty-three stage II-III dogs with loco-regionally controlled CMM across the xenogeneic vaccine studies have a KM MST of 569 days. Minimal to mild pain was noted on vaccination and one dog experienced vitiligo. We have recently investigated antibody responses in dogs vaccinated with HuTyr and found 2- to 5-fold increases in circulating antibodies to human tyrosinase. CONCLUSIONS: The results of these trials demonstrate that xenogeneic DNA vaccination in CMM: (1) is safe, (2) leads to the development of anti-tyrosinase antibodies, (3) is potentially therapeutic, and (4) is an attractive candidate for further evaluation in an adjuvant, minimal residual disease Phase II setting for CMM.

Evaluation of Elispot assays: influence of method and operator on variability of results.

In this study, a comprehensive comparative analysis of different evaluation methods of Elispot plates was performed. Three investigators using three different evaluation approaches read 50 randomly selected wells at three different time points. The methods were: (1) manual evaluation using a stereomicroscope, (2) automated evaluation using an image analysis reader system with reading parameters established by each investigator, and (3) automated evaluation using a reader system with preset reading parameters using assay-specific controls. We demonstrate that manual evaluation had the highest variability both within the same method and when comparing all methods, followed by automated evaluation with investigator-dependent parameters. The variability was low only when all investigators used the same parameters established using assay-specific controls. This variability was independent of operator or spot number per well. Based on this study, recommendations for standardization and validation procedures of Elispot assay performance and evaluation procedures are presented.

Diverse roles of conserved asparagine-linked glycan sites on tyrosinase family glycoproteins.

The tyrosinase family of genes has been conserved throughout vertebrate evolution. The role of conserved N-glycan sites in sorting, stability, and activity of tyrosinase family proteins was investigated using two family members from two different species, mouse gp75/tyrosinase-related protein (TRP)-1/Tyrp1 and human tyrosinase. Potential N-linked glycosylation sites on the lumenal domains of mouse gp75/TRP-1/Tyrp1 and human tyrosinase were eliminated by site-directed mutagenesis (Asn to Gln substitutions). Our results show that selected conserved N-glycan sites on tyrosinase family members are crucial for stability in the secretory pathway and endocytic compartment and for enzymatic activity. Different glycan sites on the same tyrosinase family polypeptide can perform distinct functions, and conserved sites on tyrosinase family paralogues can perform different functions.

Immunity against cancer: lessons learned from melanoma.

Most major advances in human cancer immunology and immunotherapy have come from studies in melanoma. We are beginning to understand the immune repertoire of T cells and antibodies that are active against melanoma, with recent glimpses of the CD4(+) T cell repertoire. The view of what the immune system can see is extending to mutations and parts of the genome that are normally invisible.

Immunization of metastatic breast cancer patients with a fully synthetic globo H conjugate: a phase I trial.

The carbohydrate antigen globo H commonly found on breast cancer cells is a potential target for vaccine therapy. The objectives of this trial were to determine the toxicity and immunogenicity of three synthetic globo H-keyhole limpet hemocyanin conjugates plus the immunologic adjuvant QS-21. Twenty-seven metastatic breast cancer patients received five vaccinations each. The vaccine was well tolerated, and no definite differences were observed among the three formulations. Serologic analyses demonstrated the generation of IgM antibody titers in most patients, with minimal IgG antibody stimulation. There was significant binding of IgM antibodies to MCF-7 tumor cells in 16 patients, whereas IgG antibody reactivity was observed in a few patients. There was evidence of complement-dependent cytotoxicity in several patients. Affinity column purification supported the specificity of IgM antibodies for globo H. On the basis of these data, globo H will constitute one component of a polyvalent vaccine for evaluation in high-risk breast cancer patients.

Alternative roles for interferon-gamma in the immune response to DNA vaccines encoding related melanosomal antigens.

Tyrosinase-related proteins-1 and -2 (gp75/TRP-1 and TRP-2) are melanosomal membrane glycoproteins recognized by antibodies and T-cells from patients with melanoma. Xenogeneic DNA immunization against gp75/TRP-1 generates antibody-dependent tumor immunity and autoimmune depigmentation. In contrast xenogeneic TRP-2 DNA immunization induces immunity mediated by CD8+ T-cells. The role of IFN-gamma in the generation of tumor immunity and autoimmune depigmentation in these two models was investigated. No tumor protection and minimal depigmentation was observed after immunization with human TRP-2 DNA in mice deficient in IFN-gamma ligand. Repletion with recombinant murine IFN-gamma restored tumor immunity. Experiments using IL4 deficient mice demonstrated that tumor immunity was unaffected but that autoimmune depigmentation was potentially accelerated, consistent with down-modulation of autoimmunity against TRP-2 by IL4. In contrast, IFN-gamma was not required for the generation of immunity to gp75/TRP-1. In fact, exogenous IFN-gamma ablated autoantibody responses against gp75/TRP-1 after xenogeneic DNA immunization, consistent with a down-regulatory effect of IFN-gamma. These results show that immunity to TRP-2 following DNA immunization uses an IFN-gamma-dependent Th1 pathway, but immunity to gp75/TRP-1 is down-regulated by IFN-gamma.

Immunization with DNA coding for gp100 results in CD4 T-cell independent antitumor immunity.

BACKGROUND: Xenogeneic DNA immunization can exploit small differences in expressed protein sequence resulting in immune recognition of self-molecules. We hypothesized that immunizing mice with xenogeneic DNA coding for the human melanosomal membrane glycoprotein gp100 would overcome immune ignorance or tolerance and result in tumor immunity. We also investigated the immunologic mechanisms of the antitumor immunity. METHODS: C57BL/6 mice were immunized with DNA coding for human gp100, mouse gp100, or control vector by gene gun. After immunization, mice were challenged with a syngeneic melanoma expressing gp100, and tumor growth was analyzed. Mice deficient in major histocompatibility complex class I or class II molecules were similarly studied to assess the immunologic mechanism of the tumor protection. RESULTS: There was significant tumor protection after vaccination with xenogeneic human gp100 DNA. Class I, but not class II, major histocompatibility complex molecules were required for tumor immunity. In addition, mice immunized with human gp100 demonstrated autoimmunity manifested as coat color depigmentation. CONCLUSIONS: Immunization with xenogeneic DNA coding for the melanosomal glycoprotein gp100 results in tumor protection and autoimmune depigmentation. These results show that xenogeneic DNA vaccines can lead to cancer immunity without CD4(+) T-cell help with potential implications for rational vaccine design.

Systemic chemotherapy.

The main use of systemic chemotherapy in metastatic melanoma remains palliative. Dacarbazine (dimethyl-1-triazeno imidazole-4-carboxamide [DTIC]) is the standard chemotherapy agent for advanced disease. The combination chemotherapy and biochemotherapy regimens have achieved higher response rates, but have not led to durable remission or improved survival. The field of systemic therapy remains in need of a more effective and less toxic treatment strategy.

Evaluation of CD8(+) T-cell frequencies by the Elispot assay in healthy individuals and in patients with metastatic melanoma immunized with tyrosinase peptide.

The lack of reproducible, quantitative assays for T-cell responses has been a limitation in the development of cancer vaccines to elicit T-cell immunity. We utilized the Elispot assay, which allows a quantitative and functional assessment of T cells directed against specific peptides after only brief in vitro incubations. CD8(+) T-cell reactivity was determined with an interferon (IFN)-gamma Elispot assay detecting T cells at the single cell level that secrete IFN-gamma. We studied both healthy individuals and patients with melanoma. Healthy HLA-A*0201-positive individuals showed a similar mean frequency of CD8(+) cells recognizing a tyrosinase peptide, YMDGTMSQV, when compared with melanoma patients prior to immunization. The frequencies of CD8(+) cells recognizing the tyrosinase peptide remained relatively constant over time in healthy individuals. Nine HLA-A*0201-positive patients with stage IV metastatic melanoma were immunized intradermally with the tyrosinase peptide together with the immune adjuvant QS-21 in a peptide dose escalation study with 3 patients per dose group. Two patients demonstrated a significant increase in the frequency of CD8(+) cells recognizing the tyrosinase peptide during the course of immunization, from approx. 1/16,000 CD8(+) T cells to approx. 1/4,000 in the first patient and from approx. 1/14,000 to approx. 1/2,000 in the second patient. These results demonstrate that modest expansion of peptide-specific CD8(+) T cells can be generated in vivo by immunization with peptide plus QS-21 in at least a subset of patients with melanoma.

Receptor-mediated uptake of antigen/heat shock protein complexes results in major histocompatibility complex class I antigen presentation via two distinct processing pathways.

Heat shock proteins (HSPs) derived from tumors or virally infected cells can stimulate antigen-specific CD8(+) T cell responses in vitro and in vivo. Although this antigenicity is known to arise from HSP-associated peptides presented to the immune system by major histocompatibility complex (MHC) class I molecules, the cell biology underlying this presentation process remains poorly understood. Here we show that HSP 70 binds to the surface of antigen presenting cells by a mechanism with the characteristics of a saturable receptor system. After this membrane interaction, processing and MHC class I presentation of the HSP-associated antigen can occur via either a cytosolic (transporter associated with antigen processing [TAP] and proteasome-dependent) or an endosomal (TAP and proteasome-independent) route, with the preferred pathway determined by the sequence context of the optimal antigenic peptide within the HSP-associated material. These findings not only characterize two highly efficient, specific pathways leading to the conversion of HSP-associated antigens into ligands for CD8(+) T cells, they also imply the existence of a mechanism for receptor-facilitated transmembrane transport of HSP or HSP-associated ligands from the plasma membrane or lumen of endosomes into the cytosol.

Induction of cellular immunity by immunization with novel hybrid peptides complexed to heat shock protein 70.

Heat shock proteins 70 (hsp70) derived from tissues and cells can elicit cytotoxic T lymphocyte (CTL) responses against peptides bound to hsp70. However, peptides can markedly differ in their affinity for hsp, and this potentially limits the repertoire of peptides available to induce CTL by the hsp immunization. Hybrid peptides consisting of a high-affinity ligand for the peptide-binding site of hsp70 joined to T cell epitopes by a glycine-serine-glycine linker were constructed. Immunization with hybrid peptides complexed to mouse hsp70 effectively primed specific CTL responses in mice and were more potent than T cell peptide epitopes alone with hsp70. In vivo immunization with hsp70 and hybrid peptides led to rejection of tumors expressing antigen with greater efficacy than immunization with peptide epitope plus hsp70. Induction of CTL responses occurred independently of CD4(+) T cells, suggesting that immunization directly primed antigen-presenting cells to elicit CD8(+) cytotoxic T cell responses without T cell help. Both peptide/hsp70 complexes and mouse hsp70 alone were able to induce cultures of mouse bone marrow-derived dendritic cells (DC) to release cytokines, including DC from endotoxin-resistant C57BL/10Sc mice. Thus, hsp70/hybrid peptide complexes can activate DC for cytokine release, providing a potential adjuvant effect that could bypass T cell help.

Insect cells as HLA-restricted antigen-presenting cells for the IFN-gamma elispot assay.

Measurement of specific cellular immune responses in patients undergoing immunotherapy is difficult. Established approaches, including cytotoxicity (e.g., 51Cr release) and cytokine release assays, require in vitro culturing for several weeks or more of patients' peripheral blood mononuclear cells (PBMC) and the addition of exogenous cytokines. Therefore, the immunological response does not reflect in vivo conditions. To address these disadvantages, we have used an interferon-gamma (IFN-gamma) Elispot assay for detecting peptide-specific CD8(+) lymphocytes in PBMC. A limitation of this assay is the lack of a reproducible source of antigen-presenting cells (APCs). Currently available APCs often lead to significant background levels. It has been shown that transfected insect cells can express empty MHC class I molecules on their surface. We have transfected Drosophila melanogaster S2 cells and the Lepidopteran line Sf9 with the gene encoding human HLA-A2.1. We demonstrate that insect cells expressing a human HLA molecule effectively function as APCs in the IFN-gamma Elispot assay. Initially the feasibility of the assay was assessed using CD8(+) T cells from HLA-A2.1(+) donors with known reactivity against an HLA-A2.1-binding epitope of the influenza matrix protein. Use of insect cells as APCs abrogated background spots, increasing sensitivity. We further observed that a short-term prestimulation of PBMC with peptide-pulsed insect cells markedly enhanced the frequency of peptide-specific T cells that could be measured in the Elispot assay without increasing the background. This approach was then used to measure CD8(+) T cell reactivity to a peptide from tyrosinase, an antigen that is processed and presented by melanoma cells. Insect cells expressing human HLA molecules provide a standard APC for monitoring CD8(+) T cell responses to tumor and viral peptides during immunotherapy.

Retrovirally transduced mouse dendritic cells require CD4+ T cell help to elicit antitumor immunity: implications for the clinical use of dendritic cells.

Presentation of MHC class I-restricted peptides by dendritic cells (DCs) can elicit vigorous antigen-specific CTL responses in vivo. It is well established, however, that T cell help can augment CTL function, raising the question of how best to present tumor-associated MHC class I epitopes to induce effective tumor immunity. To this end, we have examined the role of MHC class II peptide-complexes present on the immunizing DCs in a murine melanoma model. To present MHC class I- and II-restricted Ags reliably on the same cell, we retrovirally transduced bone marrow-derived DCs with the model Ag OVA encoding well-defined class I- and II-restricted epitopes. The importance of CD4+ T cells activated by the immunizing DCs in this model is demonstrated by the following findings: 1) transduced DCs presenting class I and class II epitopes are more efficient than class I peptide-pulsed DCs; 2) MHC class II-deficient DCs fail to induce tumor protection; 3) CD4+ T cell depletion abolishes induction of tumor protection; and 4) DCs presenting bovine serum Ags are more effective in establishing tumor immunity than DCs cultured in syngeneic serum. When MHC class II-deficient DCs were directly activated via their CD40 receptor, we indeed observed a moderate elevation of OVA-specific CTL activity. However, this increase in CTL activity was not sufficient to induce in vivo tumor rejection. Thus, our results demonstrate the potency of genetically modified DCs that express both MHC class I and II epitopes, but caution against the use of DCs presenting only the former.

Paucity of functional T-cell memory to melanoma antigens in healthy donors and melanoma patients.

The functional characteristics of CD8+ T cells specific for melanoma antigens (MAs) have often been defined after in vitro culture using nonprofessional antigen-presenting cells. We have examined CD8+ T-cell immunity to MAs and a viral antigen (influenza) in uncultured T cells of healthy donors and melanoma patients using autologous, mature, monocyte-derived dendritic cells (DCs) pulsed with peptide antigens and viral vectors. Antigen-specific IFN-gamma-producing T cells reactive with HLA-A*0201-restricted peptides from four melanoma antigens (MelanA/MART-1, MAGE-3, tyrosinase, and gp100) were detected only at low frequencies (<30 per 2 x 10(5) peripheral blood mononuclear cells for each of the MAs) from HLA-A2.1-positive healthy donors (n = 12) and patients with stages III/IV melanoma (n = 8). Detection of MA-specific, but not influenza matrix peptide (Flu-MP)-specific, T cells required a high concentration (10 microg/ml) of the peptide in this assay. Furthermore, these T cells did not recognize endogenously processed antigen on tumor cell lines or cells infected with viral vectors capable of expressing MAs. The use of autologous, mature DCs led to a significant increase in the number of Flu-MP, but not MA-specific, T cells in 16-h ELISPOT assays for both melanoma patients and healthy donors. In 1-week cocultures with DCs pulsed with 10 microg/ml peptide, MelanA/MART-1-specific T cells did not readily proliferate or differentiate into lytic effectors, in contrast to strong influenza-specific lytic responses. Therefore, despite distinct memory responses to influenza antigens, melanoma patients and healthy controls have a paucity of MA-reactive memory T cells, failing to rapidly generate IFN-gamma-secreting lytic effectors in short-term assays, even when stimulated by DCs.

Coupling and uncoupling of tumor immunity and autoimmunity.

Self-antigens, in the form of differentiation antigens, are commonly recognized by the immune system on melanoma and other cancers. We have shown previously that active immunization of mice against the melanocyte differentiation antigen, a tyrosinase-related protein (TRP) gp75(TRP-1) (the brown locus protein) expressed by melanomas, could induce tumor immunity and autoimmunity manifested as depigmentation. In this system, tumor immunity and autoimmunity were mediated by autoantibodies. Here, we characterize immunity against another tyrosinase family glycoprotein TRP-2 (the slaty locus protein), using the same mouse model and method of immunization. As observed previously for gp75(TRP-1), immunity was induced by DNA immunization against a xenogeneic form of TRP-2, but not against the syngeneic gene, and depended on CD4(+) cells. Immunization against TRP-2 induced autoantibodies and autoreactive cytotoxic T cells. In contrast to immunization against gp75(TRP-1), both tumor immunity and autoimmunity required CD8(+) T cells, but not antibodies. Only autoimmunity required perforin, whereas tumor immunity proceeded in the absence of perforin. Thus, immunity induced against two closely related autoantigens that are highly conserved throughout vertebrate evolution involved qualitatively different mechanisms, i.e., antibody versus CD8(+) T cell. However, both pathways led to tumor immunity and identical phenotypic manifestations of autoimmunity.

Phase III multicenter randomized trial of the Dartmouth regimen versus dacarbazine in patients with metastatic melanoma.

PURPOSE: Several single-institution phase II trials have reported that the Dartmouth regimen (dacarbazine, cisplatin, carmustine, and tamoxifen) can induce major tumor responses in 40% to 50% of stage IV melanoma patients. This study was designed to compare the overall survival time, rate of objective tumor response, and toxicity of the Dartmouth regimen with standard dacarbazine treatment in stage IV melanoma patients. PATIENTS AND METHODS: In this multicenter phase III trial, 240 patients with measurable stage IV melanoma were randomized to receive the Dartmouth regimen (dacarbazine 220 mg/m(2) and cisplatin 25 mg/m(2) days 1 to 3, carmustine 150 mg/m(2) day 1 every other cycle, and tamoxifen 10 mg orally bid) or dacarbazine 1, 000 mg/m(2). Treatment was repeated every 3 weeks. Patients were observed for tumor response, survival time, and toxicity. RESULTS: Median survival time from randomization was 7 months; 25% of the patients survived > or = 1 year. There was no difference in survival time between the two treatment arms when analyzed on an intent-to-treat basis or when only the 231 patients who were both eligible and had received treatment were considered. Tumor response was assessable in 226 patients. The response rate to dacarbazine was 10.2% compared with 18.5% for the Dartmouth regimen (P =.09). Bone marrow suppression, nausea/vomiting, and fatigue were significantly more common in the Dartmouth arm. CONCLUSION: There was no difference in survival time and only a small, statistically nonsignificant increase in tumor response for stage IV melanoma patients treated with the Dartmouth regimen compared with dacarbazine. Dacarbazine remains the reference standard treatment for stage IV melanoma.