6057 Background: There is a growing body of data indicating that conventional WHO and RECIST criteria do not optimally identify tumor responses associated with increased survival. Methods: We compared several response criteria to assess the efficacy of intratumoral adenoviral p53 gene therapy (Advexin) in a Phase 2 trial of 106 patients with recurrent SCCHN. Results: The percentage of patients with tumor responses defined by reductions in bi-dimensional tumor area by CT scans of 50%, 30%, 25%, 10% or stable disease of > 3 months were 8%, 10%, 11%, 16% and 20% respectively. The median survival for the entire population was 5.9 months while the responder populations defined by tumor decreases of 50%, 30%, 25%, 10% or stable disease > 3 months had median survivals of 40.8, 17.0, 13.7, 12.3 and 11.4 months respectively. There was a statistically significant increase in median survival for each of the responder populations compared to non-responders (p < 0.0009 for each comparison by logrank test). Statistical significance was maintained for the response definitions in landmark analyses excluding patients with survivals less than 3 and 6 months respectively. Locoregional disease control (CR + PR + SD > 3 months) was a stronger predictor of survival by Cox Proportional Hazard analysis (p = 0.0002) than conventional > 50% tumor reduction CR + PR response criteria (p = 0.005). With respect to abnormal p53 detected by immunohistochemistry (≥20% positive cells), 75% of patients (12/16) with p53+ tumors demonstrated locoregional disease control compared to only 18% (2/11) with p53− tumors p=0.0063 Fisher’s Exact Test. In addition, the median survival of patients with abnormal p53 was 11.6 months compared to only 3.5 months in patients with normal p53 (p=0.0007 Logrank Test). Conclusions: Our findings indicate that abnormal p53 is a predictive biomarker that identifies a subset of patients most likely to benefit from p53 gene therapy and that recurrent SCCHN tumor response definitions based upon smaller reductions in tumor size or the absence of progression (stable disease > 3 months) more accurately identified Advexin treated patients with increased survival than conventional response criteria. [Table: see text]

This review summarizes clinical experience with infectious disease vaccines and data from animal tumor models that support a paradigm shift for cancer vaccines from therapeutic to prevention applications.

The purpose of this study was to determine the safety, toxicity, and antitumor immune response following S.C. immunizations with a mixture of irradiated, autologous tumor cells and autologous fibroblasts that were genetically modified to express the gene for interleukin 2 (IL-2) in patients with colorectal carcinoma. Ten patients were treated with a fixed dose of tumor cells (10(7)) and escalating doses of fibroblasts secreting IL-2 (per 24 h): 100 units (three patients), 200 units (three patients), 400 units (three patients), and 800 units (one patient). Pre- and posttreatment peripheral blood mononuclear cells were evaluated for evidence of antitumor immune responses. Fatigue and/or flu-like symptoms were experienced by seven patients and delayed-type hypersensitivity-like skin reactions were observed at the sites of the second or subsequent vaccinations in five patients. Low frequencies of tumor cytotoxic T-cell precursors (range, 1/190,000-1/1,320,000 peripheral blood mononuclear cells) were detected prior to therapy in four of seven patients. There was a 5-fold increase following treatment in the frequency of tumor cytotoxic T-cell precursors in two of six evaluable patients. Some patients with colorectal cancer have low frequencies of tumor cytotoxic T-cell precursors that may be increased by this well-tolerated form of IL-2 gene therapy, which warrants continued clinical evaluation.

PURPOSE We have completed a phase I study, followed by three phase I/II studies, in patients with metastatic melanoma, renal cell carcinoma (RCC), and sarcoma in order to evaluate the safety, toxicity, and antitumor activity of Leuvectin (Vical Inc, San Diego, CA), a gene transfer product containing a plasmid encoding human interleukin (IL)-2 formulated with the cationic lipid 1, 2-dimyristyloxypropyl-3-dimethyl-hydroxyethyl ammonium bromide/dioleyl-phosphatidyl-ethanolamine (DMRIE/DOPE) and administered intratumorally. PATIENTS AND METHODS Twenty-four patients were treated in the phase I study. Leuvectin doses were 10 microg, 30 microg, or 300 microg weekly for 6 weeks. In three subsequent phase I/II studies, a total of 52 patients (18 with melanoma, 17 with RCC, and 17 with sarcoma) were treated with further escalating doses of Leuvectin: 300 microg twice a week for 3 weeks, 750 microg weekly for 6 weeks, and 1,500 microg weekly for 6 weeks. RESULTS There were no drug-related grade 4 toxicities and only one grade 3 toxicity, but the majority of patients experienced mild constitutional symptoms after treatment. In the phase I/II studies, 45 patients were assessable for response (14 with RCC, 16 with melanoma, and 15 with sarcoma). Two patients with RCC and one with melanoma have achieved partial responses lasting from 16 to 19 months and continuing. In addition, two RCC, three melanoma, and six sarcoma patients had stable disease lasting from 3 to 18 months and continuing. The plasmid was detected by polymerase chain reaction assay in the posttreatment samples of 29 of 46 evaluated patients. Immunohistochemistry studies on serial biopsy specimens showed increased IL-2 expression and CD8(+) infiltration after treatment in the tumor samples of several patients (12 and 16, respectively). CONCLUSION Direct intratumoral injection of Leuvectin is a safe and possibly effective immunotherapeutic approach in the treatment of certain tumor types.

The murine 96.5 monoclonal antimelanoma antibody (MoAb) was labeled with In-111, and 1-20 mg were administered to 21 patients who had proved or suspected melanoma metastases. One patient was studied twice. In four patients, unlabeled 96.5 MoAb was administered prior to the radiopharmaceutical. All of the patients tolerated the procedure without toxicity regardless of the mass of MoAb administered. The scans were interpreted by two observers, one with full knowledge, the other with no knowledge of the cases. Increasing the MoAb mass or preinfusing unlabeled MoAb prior to the administration of In-111 MoAb resulted in a prolongation of the serum half time, and appeared to improve tumor detection. Lesions were best seen at 72 hours after infusion or later. In all patients who had metastatic disease, at least one tumor site was apparent. Fifty-six per cent of known lesions 1.5 cm or greater in size were detected by the physician who had knowledge of the cases when data from all doses were considered. There were eight lesions detected that were not suspected in the workup of the patient. When these are included, the detection rate rises to 61%. Forty-nine per cent were detected by the other physician. Subtraction techniques were not employed. Lesions were often better seen with single photon emission computed tomography than with planar imaging techniques. The 96.5 In-111 MoAb appears to have utility for the detection of metastatic melanoma. Further clinical evaluation of 96.5 In-111 MoAb is warranted.

Characterization of several high-affinity murine monoclonal anticarcinoembryonic antigen (CEA) antibodies suggested good specificity except for cross-reactivity with an antigen on granulocytes and erythrocytes which was different from the previously described normal cross-reacting antigen of granulocytes. In vivo studies in athymic mice using an indium conjugate of an anti-CEA monoclonal antibody (MoAb) revealed excellent specific uptake in colorectal carcinoma xenografts. Studies were conducted in humans to determine the limitations produced by the cross-reactivity with granulocytes and erythrocytes. Patients with metastatic colorectal cancer received 3 to 6 mg of anti-CEA MoAb over 10 min or 2 hr. In five of six trials, the MoAb infusion was associated with a 40 to 90% decrease in circulating granulocytes and systemic toxicity including fever, rigors, and emesis. One patient had no change in cell count and had no toxicity. Radionuclide scans with 111In-anti-CEA MoAb showed marked uptake in the spleen when cells were eliminated, and in the liver, especially when pretreatment CEA levels were high. Metastatic tumor sites failed to concentrate the isotope. This study emphasizes the potential limitations for radioimmunodetection and/or radioimmunotherapy imposed by reactivity with circulating cells, and suggests that certain toxic reactions associated with MoAb infusions are related to destruction of circulating cells rather than allergic reactions to mouse protein. It also emphasizes how variables such as dose and binding affinity of antibody, radioisotope used, and assessment at different observation points can obscure lack of antibody specificity.

Peripheral blood or bone marrow lymphoid cells from 81 cases of chronic lymphocytic leukemia (CLL) and from 34 children and 31 adults with acute lymphocytic leukemia (ALL), were evaluated for surface immunoglobulin (sIg), cytoplasmic immunoblobulin (cIg) and with anti-la and anti-pan T-cell (T101) monoclonal antibodies (MoAbs). In childhood ALL, the majority of cases (65%) were Ia+T101-cIg- sIg-. Anti-Ia and T101 reactivity were mutually exclusive and the Ia+T101- phenotype was associated with better prognostic factors and longer median disease free survival than the Ia-Tl01+ phenotype. In adult ALL, the majority of cases (65%) were also Ia+T101-cIg-sIg-. Anti-Ia and T101 reactivity were not mutually exclusive, however, all T101+ patients presented with leukemic cell counts greater than 75, 000/mm3. In CLL, all cases were Ia+, and the most common phenotype was Ia+T101+sIg+. Paraproteinemia was present in 8/9 cases with the Ia+T101-sIg+ phenotype and proteinuria was found in 3/5 patients with the E-rosette negative, Ia+T101+sIg- phenotype. The data suggests that these MoAbs may be useful in identifying clinically relevant subgroups of acute and chronic lymphocytic leukemia.