The performance of BRCA1 immunohistochemistry for detecting germline, somatic, and epigenetic BRCA1 loss in high-grade serous ovarian cancer.

BACKGROUND: BRCA1 expression can be lost by a variety of mechanisms including germline or somatic mutation and promotor hypermethylation. Given the potential importance of BRCA1 loss as a predictive and prognostic biomarker in high-grade serous ovarian cancer, we sought to evaluate the utility of BRCA1 immunohistochemistry (IHC) in screening for BRCA1 loss by germline, somatic, and epigenetic mechanisms. PATIENTS AND METHODS: Patients with advanced high-grade serous ovarian cancer who had previously undergone germline BRCA1 testing were identified. Samples from each tumor were stained for BRCA1 and reviewed independently by two pathologists blinded to BRCA status. Tumors with abnormal BRCA1 IHC and wild-type germline testing underwent further evaluation for somatic BRCA1 mutations and promoter hypermethylation. McNemar's test was used to determine the association of BRCA1 IHC with germline BRCA1 mutations and BRCA1 loss through any mechanism. Kaplan-Meier methods were used to estimate overall survival (OS), and the log-rank test was used to assess differences between groups. RESULTS: Inter-rater reliability between the two pathologists on BRCA IHC interpretation was very good (kappa coefficient 0.865, P = 0.16; McNemar's test). BRCA1 IHC was abnormal in 36% (48/135) of cases. When compared with germline BRCA1 status, BRCA1 IHC had a high negative predictive value (95.4%) but a low positive predictive value (PPV, 52.1%). When accounting for promoter hypermethylation and somatic mutations as alternative methods of BRCA1 loss, the PPV rose to 87.5%. Five-year OS rate was 49.6% [95% confidence interval (CI) 26.3% to 69.3%] for patients with germline BRCA1 mutations, 50.4% (95% CI 27.5% to 69.5%) for germline wild-type BRCA1 and abnormal IHC, and 52.1% (95% CI 38.4% to 64.2%) for germline wild-type BRCA1 and normal IHC (P = 0.92). CONCLUSIONS: BRCA1 IHC interpretation was a highly reproducible and accurate modality for detecting germline, somatic, or epigenetic mechanisms of BRCA1 loss. These results support further development of BRCA1 IHC as a potential biomarker for BRCA1 loss in high-grade serous ovarian cancer.

Survival in epithelial ovarian cancer: a multivariate analysis incorporating BRCA mutation status and platinum sensitivity.

BACKGROUND: Patients with BRCA-associated ovarian cancer (OC) have a survival advantage over those with sporadic OC. To further explore this, we examined the impact of prognostic factors on disease-free survival (DFS) and overall survival (OS) in patients with known BRCA mutation status. PATIENTS AND METHODS: We reviewed stage III-IV OC patients treated at our institution between 1 December 1996 and 30 September 2006 and also tested on protocol for BRCA mutations. Impact on DFS and OS was determined by Kaplan-Meier analysis and a Cox proportional hazards model. RESULTS: Of the 110 patients, 36 had deleterious BRCA mutations [BRCA (+)] and 74 were BRCA wild type [BRCA(-)]. Thirty-one of 36 (86%) BRCA (+) and 60 of 74 (81%) BRCA (-) patients were platinum sensitive (P = 0.60). Median OS was longer for BRCA (+) patients (not reached versus 67.8 months; P = 0.02), but DFS was similar (26.9 versus 24.0, P = 0.3). On multivariate analysis, OS correlated with primary platinum sensitivity [HR = 0.15; 95% CI (confidence interval) 0.06-0.34] and BRCA (+) mutation status (HR = 0.33; 95% CI 0.12-0.86). CONCLUSIONS: BRCA mutation status predicted OS independent of primary platinum sensitivity, suggesting that underlying tumor biology contributes to disease outcome and may be worthy of consideration in future clinical trial design.

Circulating interleukin 6 during a continuous infusion of tumor necrosis factor and interferon gamma.

Plasma samples obtained from patients receiving a 24-h continuous infusion of human recombinant (hr)TNF or a combination of two overlapping, 24-h continuous infusions of hrIFN-gamma and hrTNF were analyzed for IL-6 in a sensitive bioassay. A transient appearance of circulating IL-6 was observed with peak levels between 3 and 6 h after the start of the hrTNF infusion. These peak levels correlated quite well with the dose of hrTNF administered (r = 0.86; p less than 0.001). The maximal value observed was 27.5 ng/ml IL-6 in a sample of a patient receiving 545 micrograms/m2 hrTNF. The combination of hrIFN-gamma (200 micrograms/m2) and hrTNF in the infusions resulted in higher IL-6 levels than a comparable dose of hrTNF alone. A maximal value of 23.5 ng/ml IL-6 was observed in a patient receiving 205 micrograms/m2 hrTNF. No IL-6 was found in the plasma of patients during the 12-h infusion with hrIFN-gamma alone, except for two borderline samples.

Identification of a hemagglutinin-specific idiotype associated with reovirus recognition shared by lymphoid and neural cells.

A xenogeneic antiserum raised to antireovirus immunoglobulin was used to define an idiotypic determinant present on antibodies to reovirus type 3 hemagglutinin. The same idiotype was identified on nonimmune lymphoid cells and on neuronal cells that specifically bind the hemagglutinin of type 3 reovirus. This idiotypic determinant, called Id3, is shared by (a) a monoclonal antibody to the neutralization site of hemagglutinin from type 3 reovirus; (b) BALB/c serum antibodies to the hemagglutinin of reovirus type 3; (c) R1.1, a murine thymoma cell line that binds reovirus type 3; (d) primary cultures of murine neuronal cells. The presence of an idiotype shared by antihemagglutinin antibodies and by structures on nonlymphoid cells suggests a general relationship between disparate receptors that recognize a common determinant. Furthermore, this suggests a novel approach for the study of viral receptor interactions and for analysis of mechanisms of autoimmune responses.

Hemagglutinin variants of reovirus type 3 have altered central nervous system tropism.

Variants of the Dearing strain of reovirus type 3 with antigenically altered hemagglutinin proteins are much less neurovirulent than the parental virus. When injected intracerebrally into mice these variants infected a subset of the brain neurons that were infected by the parental virus. When injected intraperitoneally, the variants did not spread to the brain. These results indicate that minor modifications of the reovirus hemagglutinin dramatically alter the ability of the virus to spread into and injure the central nervous system.

Tumor necrosis factor expression in human epithelial tumor cell lines.

Tumor necrosis factor (TNF) is a monokine with in vitro cytotoxicity for some but not all tumor cells. The basis for sensitivity and resistance to the antitumor effects of this agent remains unclear. The present studies have monitored the effects of TNF on 14 epithelial tumor cell lines. Eleven of these cell lines were resistant to the growth inhibitory effects of TNF (50% inhibitory concentration greater than 1,000 U/ml). 12 of the 14 tumor cell lines has detectable levels of high affinity cell surface TNF binding sites, thus suggesting that resistance was not often due to the absence of cell surface TNF receptors. Northern blot analysis demonstrated that three of the eleven resistant cell lines expressed detectable levels of TNF mRNA. Furthermore, both sensitive and resistant epithelial tumor cells had the capacity to express TNF transcripts in the presence of the protein synthesis inhibitor, cycloheximide. Finally, the presence of TNF expression at the RNA level is shown to be associated with the production of a TNF-like protein in the resistant Ov-D ovarian carcinoma cells. These findings suggest that certain human epithelial tumor cell lines inherently resistant to TNF also express this cytokine.

Specific plasma membrane receptors for reovirus on rat pituitary cells in culture.

Specific cellular and host tropism is a characteristic property of many viruses mediated by the interaction of viral attachment proteins with components of the plasma membrane of the cell. We have studied the binding of virus to cells quantitatively by using type 3 reovirus labeled with 125I and GH4C1 pituitary cells in culture. Binding was rapid at both 4 degrees and 15 degrees C and was stable over a 9-h period. Unlabeled virus inhibited binding of the labeled virus in a dose-dependent manner. Scatchard analysis revealed 4,200 viral binding sites/cell with an apparent affinity of 1.2 X 10(-11) M. Also, binding of type 3 reovirus was inhibited by antibodies directed against the viral hemagglutinin and partially inhibited by type 2 reovirus, but was unaffected by type 1 reovirus or a variety of other ligands that bind to receptors on GH4C1 cells. These data indicate that reovirus binds to a high affinity, specific receptor on target cells, which may control its tropism and ultimate disease expression.

Human granulocyte-macrophage colony-stimulating factor induces expression of the tumor necrosis factor gene by the U937 cell line and by normal human monocytes.

Human granulocyte-macrophage colony-stimulating factor (GM-CSF) exerts profound effects on the proliferation, differentiation, and effector function of myeloid lineage cells. In contrast to its growth-promoting effects on normal myeloid progenitor cells, we found that GM-CSF unexpectedly inhibited the colony growth of U937 cells in agar culture. Furthermore, medium conditioned by recombinant GM-CSF(rGM-CSF)-treated U937 cells was found to exert an inhibitory effect on subsequent U937 colony growth that was partially due to the presence of tumor necrosis factor (TNF). By Northern blot analysis, rGM-CSF was shown to induce expression of the TNF gene in U937 cells and in T-lymphocyte-depleted, monocyte-enriched peripheral blood mononuclear cells. Furthermore, rGM-CSF was observed to significantly enhance TNF secretion by monocytes stimulated with endotoxin and phorbol myristate acetate (PMA). These data suggest that some of the biological effects of GM-CSF may be amplified through the release of monokines such as TNF.

Phase I trial of bortezomib and carboplatin in recurrent ovarian or primary peritoneal cancer.

PURPOSE: To determine the maximum-tolerated dose, pharmacodynamics, and safety of the combination of bortezomib and carboplatin in recurrent ovarian cancer. PATIENTS AND METHODS: Fifteen patients were treated with a fixed dose of carboplatin (area under the curve [AUC] 5) and increasing doses of bortezomib (0.75, 1, 1.3, and 1.5 mg/m2/dose). Patients must have received upfront chemotherapy and up to two prior chemotherapy regimens for recurrent disease. Neurologic evaluation was performed at baseline and after every two cycles by the Functional Assessment of Cancer Therapy/Gynecologic Oncology Group neurotoxicity questionnaire and examination by an attending neurologist. All patients received carboplatin alone in cycle 1 to establish baseline pharmacodynamics for nuclear factor-kappa B (NF-kB). Starting with cycle 2, patients were treated with carboplatin on day 1 and bortezomib on days 1, 4, 8, and 11. RESULTS: Diarrhea, rash, neuropathy, and constipation (with colonic wall thickening on computed tomography) were dose-limiting toxicities, occurring in the two patients treated at the 1.5 mg/m2/dose level. The Functional Assessment of Cancer Therapy/Gynecologic Oncology Group neurotoxicity questionnaire was helpful in guiding the need for dose reductions. Neurotoxicity was manageable through six cycles, with appropriate dose reductions. Carboplatin had no effect on bortezomib pharmacodynamics as measured by percent inhibition of the 20S proteasome. Bortezomib decreased carboplatin-induced NF-kB. The overall response rate to this combination was 47%, with two complete responses (CR) and five partial responses, including one CR in a patient with platinum-resistant disease. CONCLUSION: The recommended phase II dose of bortezomib administered in combination with carboplatin (AUC 5) is 1.3 mg/m2/dose.

Recombinant human tumor necrosis factor administered as a 24-hour intravenous infusion. A phase I and pharmacologic study.

Recombinant human tumor necrosis factor (rH-TNF) is a cytokine with direct antitumor properties. In a phase I trial we continuously infused rH-TNF for 24 hours. We gave a total of 115 courses of therapy to 50 patients. Doses ranged from 4.5 to 645 micrograms of rH-TNF/m2. Systemic toxicity, including fever, chills, fatigue, and hypotension, increased with the dose of rH-TNF administered. Doses greater than 454 micrograms/m2 frequently caused severe lethargy and fatigue, which precluded hospital discharge of the patient at the completion of therapy. The dose-limiting toxicity was hypotension, and five patients treated at the two highest dose levels required dopamine treatment. Other organ-specific toxicity was modest and spontaneously resolved after 48 hours. The 24-hour infusions of rH-TNF were associated with significant decreases in serum cholesterol and high-density lipoprotein levels. Pharmacokinetic studies using an enzyme-linked immunosorbent assay demonstrated peak plasma rH-TNF levels of 90-900 pg/mL. Despite continuous infusion of rH-TNF, no steady-state level was achieved. The recommended phase II dose for rH-TNF as a 24-hour continuous infusion is 545 micrograms/m2.

Effects of tiazofurin on protooncogene expression during HL-60 cell differentiation.

The synthetic nucleoside analogue, tiazofurin (2-beta-D-ribofuranosylthiazole-4-carboxamide, NSC 286193) is an inhibitor of the enzyme inosine monophosphate (IMP) dehydrogenase and depletes guanine nucleotide pools. In the present study, we have monitored the effects of tiazofurin on human HL-60 promyelocytic cell differentiation and protooncogene expression. Tiazofurin (10 microM) induced a more differentiated HL-60 cell phenotype as determined by histochemical staining and decreased myeloperoxidase gene expression. This induction of differentiation was associated with a loss of proliferative capacity and decreases in clonogenic survival. The results also demonstrate that tiazofurin induces a down-regulation of c-myc mRNA levels. In contrast, there was no detectable change in the level of 3.8-kilobase c-myb transcripts. Furthermore, treatment of HL-60 cells with tiazofurin resulted in the appearance of an additional c-myb mRNA with an apparent size of 3.3 kilobases. The addition of guanosine to tiazofurin-treated HL-60 cells prevented the down-regulation of c-myc transcripts and also inhibited induction of the 3.3-kilobase c-myb transcript. Moreover, this additional transcript was not detected during induction of HL-60 cells by dimethyl sulfoxide, tumor necrosis factor, and retinal, but was induced by another IMP dehydrogenase inhibitor, mycophenolic acid. These results suggest a role for guanosine ribonucleotides in the regulation of c-myc and c-myb gene expression during HL-60 cell differentiation. The results also suggest that changes in c-myb expression can be dissociated from that of c-myc and induction of myeloid differentiation.

Posttranscriptional regulation of protein expression in human epithelial carcinoma cells by adenine-uridine-rich elements in the 3'-untranslated region of tumor necrosis factor-alpha messenger RNA.

Eukaryotic mRNAs contain 3'-untranslated regions (UTR) that are involved in posttranscriptional control of gene expression. AU-rich octanucleotide repeats, UUAUUUAU, present in the 3'-UTR of mature lymphokine and other cytokine transcripts, have been implicated in the regulation of mRNA stability and translational efficiency. For example, previous evidence suggests that the AU-rich element (ARE) present in the 3'-UTR of murine tumor necrosis factor-alpha (TNF-alpha) can affect the posttranscriptional regulation of murine TNF-alpha gene expression in hematopoietic cells. Although cytokines are produced in epithelial cells, little is known about the regulation of TNF-alpha and other cytokine gene expression by 3'-UTR elements in human malignant epithelial cells. To better understand the function of the 3'-UTR of the human TNF-alpha gene in the regulation of TNF-alpha protein production in human epithelial cancer cells, a series of luciferase reporter constructs with portions of the 3'-UTR of human TNF-alpha was transfected into human breast carcinoma cell lines ZR-75-1 and ZR-75-1R (which overexpresses TNF-alpha). The 3'-UTR of TNF-alpha markedly suppressed luciferase activity in both cell lines, and the suppression of activity was reversed by deletion of the AU-rich sequences. This suppression was quantitative, with six repeats causing more inhibition than two repeats. Increased levels of luciferase activity were observed 3 h after TNF-alpha stimulation in ZR-75-1 cells transfected by constructs containing AU-rich repeats. In addition, cytoplasmic extracts from both cell lines were assayed for factors that bind to the 3'-UTR of human TNF-alpha mRNA. RNA-protein binding activities were found in both cell lines. Competition studies showed that these proteins specifically bound to AU-rich repeats present in the 3'-UTR of TNF-alpha. No binding activity was observed when the AU-rich repeats were deleted. TNF-alpha exposure markedly increased activity of several RNA-binding proteins, especially a novel Mr 50,000-55,000 RNA-binding protein. The binding activity in untreated ZR-75-1R was higher than that in untreated ZR-75-1 cells, suggesting that the level of RNA-protein binding correlates with the expression level of TNF-alpha in human epithelial cancer cells and that the RNA-binding proteins may control expression of TNF-alpha in ZR-75-1 cells. We conclude that the AU-rich repeats in the 3'-UTR of human TNF-alpha mRNA may regulate gene expression in human epithelial cancer cells by binding to AU sequence-specific proteins, including a previously undescribed Mr 50,000-55,000 protein not observed in hematopoietic cells.

Inhibition of murine erythroleukemia cell differentiation by 3-deazaadenosine.

Recent studies have demonstrated that 5'-methylthioadenosine, an inhibitor of S-adenosylhomocysteine (AdoHcy) hydrolase, blocks induction of murine erythroleukemia cell (MEL) differentiation. The nucleoside analogue 3-deazaadenosine (c3Ado) is both an efficient substrate and a potent inhibitor of AdoHcy hydrolase. The present study was undertaken to determine whether c3Ado would similarly inhibit MEL differentiation. The results demonstrate that c3Ado inhibits induction of MEL differentiation by dimethyl sulfoxide, hexamethylene bisacetamide, butyric acid, and diazapam. c3Ado blocks the appearance of the differentiated MEL phenotype by inhibiting both MEL heme synthesis and transcription of alpha- and beta-globin RNA. The inhibitory effect of c3Ado on MEL differentiation is concentration dependent, reversible, and potentiated by L-homocysteine thiolactone. Furthermore the AdoHcy/AdoMet ratio increases nearly 3.5-fold after 24 h of treatment with 50 microM c3Ado. In contrast, this c3Ado effect is not associated with polyamine depletion or cytostasis. These findings indicate that c3Ado blocks the induction of MEL differentiation at a transcriptional level and that this effect may be related to inhibition of AdoHcy hydrolase.

Continuous infusion of high-dose 1-beta-D-arabinofuranosylcytosine: a phase I and pharmacological study.

High doses of 1-beta-D-arabinofuranosylcytosine, administered as a continuous i.v. infusion, were evaluated in a Phase I trial in 14 patients with advanced solid tumors. 1-beta-D-Arabinofuranosylcytosine was given at 250 mg/sq m/h for infusions of 12 to 36 h. The mean steady state 1-beta-D-arabinofuranosylcytosine plasma level was 19.6 microM with a range of 9 to 59 microM. The principal toxicity was myelosuppression. An infusion of 18 h was well tolerated by most patients. A Phase II dose of 250 mg/sq m/h for 24 h can be used if platelet support is available. This dose schedule may be useful in the treatment of hematological disorders or in clinical combinations with DNA-damaging agents in the treatment of solid tumors.

BRCA1 up-regulation is associated with repair-mediated resistance to cis-diamminedichloroplatinum(II).

We sought to identify novel genes associated with cis-diamminedichloroplatinum(II) (CDDP) resistance, and by differential display analysis, we found that the human breast and ovarian cancer susceptibility gene BRCA1 was overexpressed in CDDP-resistant MCF-7 cells. A recent report that BRCA1 and human Rad51 colocalize in S-phase cells suggests a role for BRCA1 in DNA damage repair. We hypothesized that BRCA1 plays a role in DNA damage repair-mediated CDDP resistance. In CCDP-resistant variants of breast and ovarian carcinoma cell lines, MCF-7 CDDP/R and SKOV-3 CDDP/R, we found increased levels of BRCA1 protein, and we determined that the SKOV-3 CDDP/R cell line is significantly more proficient at DNA damage repair. Antisense inhibition of BRCA1 in this cell line resulted in an increased sensitivity to CDDP, a decreased proficiency of DNA repair, and an enhanced rate of apoptosis. These data support the hypothesis that BRCA1 is a gene involved in DNA damage repair.

Phospholipase A2 activation and autoinduction of tumor necrosis factor gene expression by tumor necrosis factor.

Tumor necrosis factor (TNF) acts via a cell surface receptor to induce a variety of cellular events including cytolysis, differentiation, and mitogenesis. The mechanisms underlying the cell specific actions of TNF are not known. In the present study, postreceptor events associated with the autoinduction of TNF expression were examined in HL-60 cells. There was no detectable alteration in phospholipase C activity as measured by inositol phosphate generation or release of choline metabolites following TNF stimulation. However, TNF increased the release of arachidonic acid metabolites from HL-60 cells. This increase in arachidonic acid metabolism was associated with a 40% increase in phospholipase A2 activity. Furthermore, the release of arachidonic acid metabolites was blocked by inhibitors of phospholipase A2. Taken together, these findings indicated that TNF stimulates phospholipase A2 and arachidonic acid metabolism in HL-60 cells. The results also demonstrate that TNF expression is induced 15-30 min after stimulation with TNF and that this effect is associated with an increase in the rate of TNF transcription. This autoinduction of TNF mRNA was blocked by inhibitors of phospholipase A2. While the cyclooxygenase inhibitor indomethacin had no detectable effect, ketoconazole and nordihydroguaiaretic acid, inhibitors of lipoxygenase, also blocked the induction of TNF expression by TNF. These findings suggest that phospholipase A2 and lipoxygenase activity are required for the transcriptional activation of TNF gene expression associated with TNF stimulation of HL-60 cells.

Fludarabine phosphate (NSC 312878) infusions for the treatment of acute leukemia: phase I and neuropathological study.

Fludarabine phosphate (NSC 312878), an adenosine deaminase resistant analogue of 9-beta-D-arabinofuranosyladenine, has entered clinical trials. Eleven patients with acute leukemia in relapse received 14 courses of fludarabine phosphate as a 5-day continuous infusion administered at doses of 40 to 100 mg/m2/day. Toxicity was characterized by uniform myelosuppression, as well as occasional nausea, vomiting, and hepatotoxicity. Three episodes of metabolic acidosis and lactic acidemia were noted. In addition, three patients suffered neurotoxicity. Two of these three patients had a severe neurotoxicity syndrome characterized by blindness, encephalopathy, and coma. Neither patient recovered neurological function. Neuropathological findings at autopsy were characterized by a diffuse, necrotizing leukoencephalopathy which was most severe in the occipital lobes. The medullary pyramids and posterior columns were also severely affected. This sporadic fatal neurotoxicity was observed only at doses greater than 40 mg/m2/day. The maximum tolerated dose for a 5-day infusion of fludarabine phosphate is thus 40 mg/m2/day.

Recombinant human tumor necrosis factor administered as a five-day continuous infusion in cancer patients: phase I toxicity and effects on lipid metabolism.

Recombinant human tumor necrosis factor (rH-TNF) is a cytotoxic monokine with pleiotropic effects. A phase I trial of rH-TNF was initiated using a five-day continuous intravenous (IV) infusion repeated every 28 days. Thirty-eight courses of therapy were administered to 19 patients. The starting dose was 5 X 10(4) U/m2/d, with escalations to 1.0 X 10(5), 2.0 X 10(5), 2.4 X 10(5), and 3.0 X 10(5) U/m2/d. Systemic side effects, including fever, chills, hypotension, fatigue, anorexia, and headaches, were mild and self-limiting. At the maximum tolerated dose of 3.0 X 10(5) U/m2/d, dose-limiting hematologic toxicity was manifested by transient thrombocytopenia and leukopenia. Elevated bilirubin levels were also seen at the higher dose levels. Lipoprotein analysis demonstrated that the five-day treatment with rH-TNF was associated with decreases in high-density lipoproteins, as well as increases in triglycerides and very-low-density lipoproteins. Pharmacokinetic studies using an enzyme-linked immunosorbent assay (ELISA) test indicated plasma rH-TNF levels less than 0.2 U/mL. The recommended phase II dose of rH-TNF administered as a five-day continuous infusion is 2.4 X 10(5) U/m2/d.

A phase I trial of recombinant human tumor necrosis factor and interferon-gamma: effects of combination cytokine administration in vivo.

The combination of tumor necrosis factor (TNF) and interferon-gamma has synergistic bioactivity in numerous preclinical model systems. We have tested this potential synergism in vivo by administration of both cytokines to patients with advanced cancer using overlapping 24-hour continuous intravenous (IV) infusions in a phase I trial. Thirty-six patients were treated with a fixed dose of interferon-gamma (200 micrograms/m2/d) with interpatient dose escalation of TNF (from 5 to 205 micrograms/m2/d). The dose-limiting toxicity at the maximal-tolerated dose (MTD) of TNF (205 micrograms/m2) with interferon-gamma was hypotension. Other toxicities noted included an influenza-like syndrome, transient decreases in circulating leukocyte and platelet counts, subclinical evidence of disseminated intravascular coagulation, and the sporadic occurrence of acute pulmonary toxicity. The recommended phase II dose for this combination schedule is TNF, 136 micrograms/m2, with interferon-gamma, 200 micrograms/m2. The addition of interferon-gamma to TNF resulted in a greater than three-fold increase in toxicity compared with TNF administered as a single agent, supporting the hypothesis that the combination of these cytokines may induce synergistic effects in vivo.