Impact of cyclophosphamide on relationships between carboplatin exposure and response or toxicity when used in the treatment of advanced ovarian cancer.

PURPOSE: To determine (1) the impact of cyclophosphamide 600 mg/m2 on previously defined relationships between carboplatin area under the plasma concentration versus time curve (AUC) and indices of toxicity and response in women with advanced ovarian cancer; and (2) the relationships between indices of cumulative drug exposure and clinical outcomes. METHODS: Carboplatin AUC = dose/(creatinine clearance [CCr] + 25) and was calculated in 224 women who received carboplatin 300 mg/m2 and cyclophosphamide 600 mg/m2. The likelihood of grade 3 or greater myelotoxicity at any carboplatin AUC was compared with the likelihood of myelotoxicity at the same single-agent carboplatin AUC. The nadir count predicted using the University of Maryland single-agent carboplatin dosing formula was compared with the nadir count observed. Received and relative-received dose-intensity were calculated. Carboplatin exposure-intensity was defined by substituting cumulative carboplatin exposure for total dose. Relationships were sought between these indices and therapeutic outcomes. RESULTS: The incidence of leukopenia and thrombocytopenia at any carboplatin AUC was greater for the two-drug combination than for single-agent carboplatin. The platelet nadir in 83% of patients was less than or equal to the nadir predicted for the same single-agent carboplatin AUC. Despite a narrow range of received dose-intensities, carboplatin exposure-intensity was distributed over a twofold range. There were no relationships between received and relative-received dose-intensity or carboplatin exposure-intensity and time to progression or survival. CONCLUSION: Any carboplatin AUC when administered with cyclophosphamide 600 mg/m2 produces greater myelotoxicity than the same AUC of single-agent carboplatin. Received carboplatin dose-intensity underestimates the range of plasma drug exposure resulting from a fixed carboplatin dosing regimen. Whether higher carboplatin exposures can improve outcome requires prospective validation.

Relationships between carboplatin exposure and tumor response and toxicity in patients with ovarian cancer.

PURPOSE: The study was undertaken to define the relationship between tumor response and carboplatin area under the curve (AUC) in patients with ovarian cancer; to study the relationship between carboplatin AUC and myelosuppression in the same population; to establish the true impact of carboplatin AUC, prior therapy, and pretreatment platelet and WBC counts on toxicity; and to define an optimal carboplatin exposure for treating patients with ovarian cancer. METHODS: With the equation AUC = dose/(glomerular filtration rate [GFR]+25), carboplatin AUC (course 1) was calculated for 1,028 patients (450 previously untreated) who received single-agent carboplatin (40 to 1,000 mg/m2) for advanced ovarian cancer. GFR was measured (chromium-51-edathamil [51Cr-EDTA] or creatinine clearance) in all patients. RESULTS: Regression analysis showed that carboplatin AUC, prior treatment, and Eastern Cooperative Oncology Group grade performance status (PS) are predictors of tumor response, thrombocytopenia, and leukopenia. Pretreatment platelet and WBC counts are additional predictors of thrombocytopenia and leukopenia, respectively. Although the likelihood of tumor response increased with increasing carboplatin AUC, this relationship was nonlinear. In all patient subsets, the likelihood of complete response (CR) or overall response did not increase significantly above a carboplatin AUC of 5 to 7 mg/mL x minutes. At any given carboplatin AUC, thrombocytopenia occurred more frequently than leukopenia, although both approached 100% as carboplatin AUC increased. Both thrombocytopenia and leukopenia were more frequent in pretreated than in untreated patients regardless of pretreatment count. At any carboplatin AUC, the influence of PS on likelihood of response and toxicity was profound. CONCLUSION: Carboplatin dosing by AUC will lead to more predictable toxicity, and increasing carboplatin AUC above 5 to 7 mg/mL x minutes does not improve the likelihood of response but does increase myelotoxicity. Therefore, careful evaluation of high-dose carboplatin therapy in a prospective, randomized trial is needed before such treatment becomes accepted practice.

Modeling toxicity and response in carboplatin-based combination chemotherapy.

Data from women with advanced ovarian cancer (International Federation of Gynecology and Obstetrics stage III or IV) were analyzed to evaluate the pharmacokinetic/pharmacodynamic relationships of carboplatin-based combination chemotherapy. With the equation area under the plasma concentration versus time curve (AUC) = dose/(creatinine clearance + 25), carboplatin AUC was calculated in each of up to six treatment cycles in 224 women with advanced ovarian cancer who had been randomized to receive carboplatin 300 mg/m2 plus cyclophosphamide 600 mg/m2. In addition, for each patient, the predicted nadir count (obtained by rearranging the University of Maryland single-agent carboplatin dosing formula) was compared with the actual observed nadir count, received and relative received dose intensities were calculated, and carboplatin exposure intensity was defined. Relationships were sought between these treatment indices and the clinical outcomes of time to progression and survival. When combined with cyclophosphamide 600 mg/m2, any carboplatin AUC was found to be associated with greater myelotoxicity and a higher likelihood of both leukopenia and thrombocytopenia occurring than had been determined for single-agent carboplatin. Furthermore, the platelet nadir in 83% of patients was equal to or below that predicted to result from the same dose of single-agent carboplatin. There was a relatively narrow range of received dose intensities within this patient population, but carboplatin exposure intensity was calculated as being distributed over a two-fold range within the population. Therefore, received carboplatin dose intensity underestimates the range of plasma drug exposure associated with a fixed dosing regimen of carboplatin. However, there were no consistent relationships between received dose intensity, relative received dose intensity, or carboplatin exposure intensity and the clinical outcomes of time to progression or survival. The relationships between carboplatin exposure and the pharmacodynamic measures of toxicity and response are likely to require definition in each regimen that includes carboplatin and for each tumor type treated.

Structural studies of orbivirus particles.

We are using crystallographic methods to investigate the structure of AHSV and BTV. Our initial approach was to investigate the structure of the major protein component of the viral core, VP7(T13). This trimeric protein has been studied in several crystal forms from both orbiviruses and reveals a structure made up of conserved domains, capable of conformational changes and possessing a cleavage site. Further crystallographic analyses of native particles have provided a picture of the VP7(T13) and VP3(T2) layers of the BTV core. The VP7(T13) layer consists of 260 trimers arranged rather symmetrically and possessing very similar structures, thereby following the rules of quasi equivalence. The VP3(T2) layer is thin and contains 120 copies of 100 kDa protein arranged as 60 approximate dimers. This type of icosahedral construction has not been observed before and appears to contain a genome which is highly ordered. We anticipate that all of these features will be common to AHSV.

VP7 from African horse sickness virus serotype 9 protects mice against a lethal, heterologous serotype challenge.

An established mouse model system was used to evaluate the effectiveness of the major outer core protein VP7 of African horse sickness virus (AHSV) serotype 9 as a subunit vaccine. Balb C mice were immunised with VP7 crystals purified from AHSV infected BHK cells. In groups of mice, each of which was immunised with > or = 1.5 micrograms of the protein in Freund's adjuvant, > or = 80% of mice survived challenge with a virulent strain of a heterologous AHSV serotype (AHSV 7), that killed > or = 80% of the mice in the uninoculated control groups. This level of protection was significantly greater than that observed in mice inoculated with equivalent amounts of either denatured VP7 (50% survival), or GST/VP7 fusion protein (50-70% survival), or which were vaccinated with AHSV 9 (40-50% survival). The VP7 protein folding, or its assembly into crystals, are thought to play some role in the effectiveness of the protective response observed. Titres of circulating antibodies against AHSV VP7 were determined by competitive ELISA but did not appear to correlate with the levels of protection observed. Passive transfer of these antibodies to syngeneic recipients also failed to protect Balb C mice from the AHSV 7 challenge. The observed protection is therefore unlikely to be due to an antibody mediated immune response.

Development of a mouse model system, coding assignments and identification of the genome segments controlling virulence of African horse sickness virus serotypes 3 and 8.

Attenuated (att) and wild type (wt) strains of the nine AHSV serotypes were evaluated for virulence in adult Balb C mice. Although most were avirulent in this system, isolates of AHSV 1att, 3wt, 3att, 4wt, 5att, 7att and 8att caused some mortality when administered via an intranasal route. After plaque cloning, only the attenuated vaccine strain of AHSV 7att caused any mortality via an intravenous route. AHSV 3att and AHSV 8wt were virulent (V) and avirulent (AV) (respectively) in the mouse model and were selected as parental strains for production of genome segment reassortants. These progeny virus strains were plaque cloned, then characterised to identify the genome segments that influence virulence of AHSV in the mouse model. Three virulence phenotypes were observed: fully virulent (V); fully avirulent (A); and a novel intermediate virulence (N) not expressed by either parental strain. Genome segment 2 (encoding outer capsid protein VP2) from the avirulent parent appeared to have a controlling influence in production of the A phenotype. Reassortants with the V phenotype all contained segment 2 from the virulent parent, however in each case they also contained genome segments 5 and 10, also from AHSV 3 (V). Genome segments 5 and 10 encode the smaller outer capsid protein VP5 and the non structural proteins NS3/NS3a, respectively. A combination of genome segments 2, 5 and 6 from the avirulent parent and segment 10 from the virulent parent were found in each of the virus strains with the N phenotype. However, comparison of two reassortants (A79 and A790), which differ only in a single segment, showed that replacement of genome segment 10 from the avirulent parent with that from the virulent parent, conferred the N phenotype on A790.

Competitive ELISA for the detection of antibodies against epizootic haemorrhagic disease of deer virus.

A competitive enzyme-linked immunosorbent assay is described for the detection of antibodies against epizootic haemorrhagic disease of deer viruses (EHDV). Test antisera were tested against a guinea-pig antiserum raised against EHDV core particles. The assay detected antibodies to all serotypes of EHDV, but unlike the agar gel immunodiffusion (AGID) test, gave no cross-reactions with antibodies against bluetongue, Palyam and Tilligery viruses. The C-ELISA would be ideal for use in epidemiological surveys since it is suitable for the examination of antisera from all susceptible species without the need for individual species-specific enzyme conjugates.

A serogroup specific enzyme-linked immunosorbent assay for the detection and identification of African horse sickness viruses.

A serogroup specific, indirect, sandwich ELISA was developed for the rapid detection of African horse sickness virus and viral antigens in field samples or in infected tissue cultures. The assay was shown to be highly sensitive and capable of providing confirmation of clinical diagnosis within one day. The results demonstrated that this ELISA will be useful for epidemiological surveillance of insect and mammalian host populations.

Monoclonal antibody based competitive ELISA for the detection of antibodies against epizootic haemorrhagic disease of deer virus.

A monoclonal antibody based competitive enzyme-linked immunosorbent assay (MC-ELISA) for the detection of antibodies against epizootic haemorrhagic disease of deer viruses (EHDV) is described. Test sera were competed with a monoclonal antibody against the VP7 protein of EHDV serotype 1. The assay was capable of detecting antibodies to all serotypes of EHDV but unlike the agar gel immunodiffusion (AGID) test, gave no cross-reactions with antibodies against bluetongue, Palyam or Tilligery viruses. The MC-ELISA was more sensitive than a polyclonal based ELISA reported previously (Thevasagayam et al., 1995b) and would be ideal for epidemiological surveys since it is suitable for the examination of antisera from all animal species without the need for individual anti-species enzyme conjugates.

Detection and differentiation of epizootic haemorrhagic disease of deer and bluetongue viruses by serogroup-specific sandwich ELISA.

A serogroup specific sandwich ELISA was developed for the detection of epizootic haemorrhagic disease of deer viruses (EHDV) in infected insects and tissue culture preparations. Polyclonal rabbit antiserum against purified EHDV core particles was used to capture viral antigen and specific binding detected using guinea pig antisera against EHDV core particles followed by anti-guinea pig immunoglobulin enzyme-labelled conjugate. The assay is EHDV specific and detects all 8 serotypes. No cross-reactions were found with related viruses such as bluetongue (BTV), Palyam, Tilligery or African horse sickness virus (AHSV). A similar serogroup specific sandwich ELISA was also developed for BTV. The assays showed a similar sensitivity in detecting the respective EHDV or BTV antigens in a pool of 500 midges where only 2 were infected. These assays allow a simple and rapid means of detecting and differentiating members of these closely related serogroups. The sensitivity of the tests will allow more extensive studies on vector competence and virus/vector distribution.

A group-specific, indirect sandwich ELISA for the detection of equine encephalosis virus antigen.

A polyclonal antibody-based, group-specific, indirect, sandwich ELISA (S-ELISA) for the detection of equine encephalosis virus (EEV) antigen was developed. Purified EEV particles were titrated in the S-ELISA and the limit of detection was determined to be approximately 9.0 ng of antigen/ml (0.45 ng/well). Positive S-ELISA reactions were recorded with seven serologically distinct EEV serotypes. No cross-reactions were recorded with other arboviruses including African horse sickness virus (AHSV) serotypes 1-9, bluetongue serotypes 1-24, epizootic haemorrhagic disease serotypes 1-8 and isolate 318, and selected isolates of Palyam, Eubenangee, Corriparta, Warrego, Akabane and bovine ephemeral fever viruses. The assay proved to be sensitive and specific for the rapid detection of EEV in cell cultures and in homogenated suckling mouse brain (MB). The data generated in this study suggest that the ELISA will be valuable for epidemiological studies of EE and will assist in making a reliable differential diagnosis between EEV and AHSV infections.

A competitive ELISA for the detection of group-specific antibody to equine encephalosis virus.

A polyclonal antibody-based, group-specific, competitive ELISA (C-ELISA) for the detection of antibodies to equine encephalosis virus (EEV) was developed. The assay measures the competition between a specific guinea pig antiserum and a test serum, for a pre-titrated EEV antigen. The C-ELISA detected antibodies to the seven known EEV serotypes. Reference antisera raised against other arboviruses did not cross react with EEV antigen. Negative sera from horses in the United Kingdom were used to establish the baseline for a negative population. Negative and positive populations of South African horses, selected on the basis of virus neutralisation were assayed subsequently. Optimal test parameters, where sensitivity≅specificity≅100%, were calculated by two-graph receiver operator characteristic (TG-ROC) analysis to be at a cut-off value of 29.5% inhibition. Results show the EEV C-ELISA described to be sensitive, specific and reliable. Used in conjunction with ELISAs available for African horse sickness virus (AHSV), differential serological diagnosis between EEV and AHSV can be achieved.

Presence of a covalently linked protein on calicivirus RNA.

The infective RNA of the calicivirus, vesicular exanthema virus, has been shown to contain a protein which is apparently linked to the RNA by a covalent bond. The protein remained bound to the RNA after boiling with SDS-mercaptoethanol-urea or treating with formamide-dimethylsulphoxide but was removed by incubating with proteinase K. The mol. wt. of the protein was estimated to be about 1o X 1O(3) by electrophoresis in highly cross-linked polyacrylamide gels. The infectivity of the RNA was destroyed by removal of the protein with proteinase K.

Purification and properties of virus particles, infectious subviral particles, and cores of bluetongue virus serotypes 1 and 4.

Effective purification methods have been developed for virus particles, infectious subviral particles (ISVP), and virus cores of bluetongue virus (BTV) serotypes 1 and 4. The purified particles were analysed by indirect ELISA or PAGE using either silver staining, or fluorography of [35S]methionine-labelled preparations. No significant contamination with host cell proteins, or with the majority of BTV nonstructural proteins was detectable in any of the particle preparations. In addition to the two major outer capsid and five core proteins previously described, the purified virus particles of both serotypes were consistently found to contain small amounts of BTV protein NS2, previously regarded as exclusively nonstructural. This protein could be removed from the particle surface by treatment with a combination of chymotrypsin and sodium N-lauroyl sarcosinate, which also resulted in the cleavage of the larger of the two major outer capsid components (protein VP2). Two of the cleavage products of VP2 and the whole of the other major outer capsid component (protein VP5) formed a modified outer capsid layer in the resultant ISVP. These subviral particles were as or more infectious than the intact virus particles but had lost haemagglutinating activity. The core-associated RNA polymerase remained inactive in ISVP.

Crystallization and preliminary X-ray analysis of the core particle of bluetongue virus.

Core particles of bluetongue virus serotype 1 (South Africa) have been crystallized. The crystals, which grow up to 0.8 mm in diameter, belong to a primitive orthorhombic space group and have point group symmetry 222. The unit cell dimensions are 754 x 796 x 823 A3 and the crystallographic asymmetric unit contains one-half of a core particle. The best crystals diffract strongly to 4.8 A Bragg spacings, which is the maximum resolution to which we can measure data with the detectors available, suggesting that useful diffraction extends well beyond this. Core particles of serotype 10 have also been crystallized but the crystals have yet to be analyzed by X-ray diffraction.

A comparison of six different bluetongue virus isolates by cross-hybridization of the dsRNA genome segments.

The relationship between six different isolates of BTV was analyzed by cross-hybridization of genomic dsRNA using blotting and probe techniques (using an alkali fragmented probe made from BTV dsRNA). The viruses compared in this way included BTV serotype 1 from South Africa, serotypes 3 and 4 from Cyprus, serotype 10 from North America, and serotypes 1 and 20 from Australia. Under the hybridization and washing conditions used, which were calculated to allow stable duplex formation between RNA molecules containing greater than 90% sequence homology, two of the genome segments (segments 2 and either 5 or 6, which encode the two major outer capsid proteins VP2 and VP5) appeared to contain serotype-specific RNA sequences. Significant cross-hybridization between these segments from different serotypes was detected only with serotypes 4 and 20, which are known to have a particularly close antigenic relationship. The amounts of homologous sequence that were detected in segments other than 2 and 5 between different viruses indicated some correlation between their geographical origins and a degree of relatedness, which is independent of the virus serotype. High levels of sequence homology were detected between the isolates from Cyprus and Africa and to a slightly lesser extent from North America, suggesting a common ancestry. These results also indicated that within the limited number of viruses studied, the Australian isolates form a separate interrelated group of bluetongue viruses.

The expressed VP4 protein of bluetongue virus binds GTP and is the candidate guanylyl transferase of the virus.

A minor core protein, VP4, of bluetongue virus serotype 10 (BTV-10) has been synthesized in insect cells infected with a genetically manipulated recombinant baculovirus. When insect cells were coinfected by this recombinant virus and a recombinant baculovirus expressing the two major core proteins (VP3 and VP7) of the virus, core-like particles (CLPs) consisting of all three proteins were formed. Purified CLPs reacted with [32P]GTP which was covalently bound to VP4 only. Similarly reconstituted CLPs with VP1 or VP6 did not form covalent complexes with [32P]GTP. The virion-derived VP4 was also shown to have GTP-binding activity. The covalent binding of GTP indicates that expressed VP4 not only is biologically active but also is the candidate guanylyl transferase of the virus. The optimum reaction conditions for GTP binding by VP4 have been investigated.

A model for vesicular exanthema virus, the prototype of the calicivirus group.

The structure of vesicular exanthema virus, the prototype member of the calicivirus group, has been studied in more detail. The RNA comprises 18% of mol. wt. of about 2.8 x 10(6), based on polyacrylamide gel electrophoresis experiments in the presence of formaldehyde. The virus contains one major polypeptide, mol. wt. 70 x 10(3) as determined by polyacrylamide gel electrophoresis and by chromatography on Sepharose 6B in the presence of 6 M-guanidine. Further evidence for the presence of a single major polypeptide was obtained by tryptic peptide analysis of 35S-methionine labelled virus. The mol. wt. of a protein oligomer produced by adjusting the pH of virus suspensions to 3.5 was c. 200 x 10(3). On the basis of these data we propose a T = 3 model for the virus capsid incorporating 180 copies of the virus protein.

Identification of a bluetongue virus serotype 1-specific ovine helper T-cell determinant in outer capsid protein VP2.

Ovine T-cell lines (including one clone [101A]), which are specific for Bluetongue virus serotype 1 (BTV1), have been established and characterized. Although these T-cell lines react with different isolates of BTV1 (including those from South Africa, Australia, Nigeria, and Cameroon), they do not react with heterologous BTV serotypes. Antigen specificity of these T-cells was studied using purified virus particles, infectious subviral particles (ISVP) and cores, or using individual BTV structural proteins that were either isolated by SDS-PAGE or expressed by recombinant strains of vaccinia virus. The results showed that each of the T-cell lines reacted with outer capsid protein VP2 (the BTV protein exhibiting most serotype-specific variation and the major neutralization antigen). However, all of the uncloned T-cell lines also reacted with either the core structural proteins or the outer capsid protein VP5. In contrast, the T-cell clone 101A only reacted with outer capsid protein VP2. Cell surface marker analysis showed that 101A has a helper T-cell phenotype (CD5+, CD4+, CD8-, T-19-). The T-cell lines and clone 101A all produced large amounts of interleukin 2 (IL-2) when stimulated with purified BTV1 virus particles, or with VP2 (up to 120 IU/ml from 2 x 10(5) T-cells). BTV serotype-specific antigenic sites, for B cells and at least one site for ovine helper T-cells, are therefore located within VP2.

The detection of African horse sickness virus antigens and antibodies in young Equidae.

Four ponies were each inoculated with a different serotype of African horse sickness virus (AHSV) which had been passaged through cell culture in order to achieve attenuation. Three of the ponies died suddenly after showing mild clinical signs, the fourth pony remained clinically normal and was killed at day 38. Infectious AHSV was isolated from blood samples collected at intervals from all four ponies. Positive antigen ELISA reactions were only observed with blood samples from two of the ponies on the two days preceding death. Specific AHSV antibodies were detected by ELISA in serum samples from the other two ponies although one eventually died. African horse sickness viral antigens were detected by ELISA in post-mortem tissue samples collected from all four ponies. No infectious virus could be detected in tissue samples taken post-mortem from the pony which survived African horse sickness (AHS) infection. In the event of a suspected outbreak of AHS it is recommended that sera and heparinized blood should be tested for specific antibodies and AHSV antigen respectively. When available, post-mortem tissues, including spleen, heart, lung and liver, should also be tested for AHSV antigen. Although the ELISA used for the detection of AHSV antigen is highly sensitive and specific, negative ELISA results should be confirmed by virus isolation attempts.