Breast cancer resistance protein expression and resistance to daunorubicin in blast cells from patients with acute myeloid leukaemia.

Breast cancer resistance protein (BCRP) is a recently described member of the ATP binding cassette transporter superfamily. It has been shown to confer resistance to mitoxantrone, topotecan, doxorubicin and daunorubicin in human tumour cell lines. We describe a study of BCRP expression in blast cells derived from 20 patients with acute myeloid leukaemia (AML). Twelve samples were from patients who had received previous cytotoxic therapy. BCRP expression was measured by immunocytochemistry using the BXP-34 monoclonal antibody. In vitro drug sensitivity was assessed using the methyl thiazol tetrazoliumbromide assay. BCRP expression varied between patients, and six out of 22 (27%) samples had > 10% cells staining positively (median 37%, range 13-95%). BCRP positivity was seen in both de novo samples and those from previously treated patients. There was a marked variation in the effect of all drugs tested between patients. Although there was no correlation between BCRP positivity and the effect of mitoxantrone, topotecan or doxorubicin, the median daunorubicin LC(50) value of BCRP(+) cells was fourfold higher than that of BCRP- cells (0.89 micromol/l compared with 0.21 micromol/l, P < 0.05). These results suggest that BCRP may be involved in resistance to the agents commonly used in AML and may explain some of the anomalous results found when studying other membrane transporters, such as P-gp or MRP.

Chemosensitivity testing predicts survival in ovarian cancer.

The aim of this study was to assess the use of the MTT assay for chemosensitity testing to identify drug resistance and predict survival in patients with advanced ovarian cancer. Samples of ascitic fluid and/or solid biopsies were taken from 120 patients with FIGO stage III or IV ovarian adenocarcinoma at presentation. Cells were exposed for 48 hours to four concentrations of clinically relevant drugs including platinums, anthracyclines and alkylating agents. Cell survival was measured using the 3-4,5-dimethyl-2, 5-diphenyl tetrazolium bromide (MTT) assay allowing patients to be grouped as "sensitive" or "resistant" in vitro. Clinical data including age, residual disease, histological grade, treatment, response after initial treatment and overall survival were collected. There was a highly significant (p<0.0001) correlation of in vitro sensitivity with in vivo response in the patients who completed their therapy, with an 83% positive predictive accuracy for resistance. This translated in the longer term to an increased survival for the patients found to be sensitive in vitro to their therapy with a 5-year survival rate of 24% compared to 12% for the resistant group (p=0.033). These results suggest that MTT chemonsensitivity testing can predict response in ovarian cancer leading to the prospect of increased survival in this devastating disease by customising therapy to individual patients.

Circumvention of ara-C resistance by aphidicolin in blast cells from patients with AML.

Treatment failure in AML is often attributed to P-glycoprotein-associated multidrug resistance. However, the importance of increased DNA repair in resistant cells is becoming more apparent. In order to investigate the ability of the DNA repair inhibitor aphidicolin to modulate drug resistance, we continually exposed blasts cells, isolated from 22 patients with AML, to a variety of agents +/- 15 microM aphidicolin for 48 hours. Cell survival was measured using the MTT assay. Overall, there was no significant effect of aphidicolin on sensitivity to daunorubicin, doxorubicin, etoposide or fludarabine. However, there was a marked increase in sensitivity to ara-C with a median 4.75-fold increase overall (range 0.8-80-fold;P< 0.005). The effect of aphidicolin was significantly greater in blast cells found resistant in vitro to ara-C (8.9-fold compared to 2.12-fold, P< 0.01). This observation was further validated by the correlation between ara-C LC(50)and extent of modulation effect (P< 0.05). Cells isolated from 10 cord blood samples were also tested in order to establish the haematological toxicity of combining ara-C and aphidicolin. The therapeutic index (LC(50)normal cells/tumour cells) for ara-C + aphidicolin was higher than that for ara-C alone suggesting no increased myelotoxicity for the combination. Increased cytotoxicity without increased haematotoxicity makes the combination of ara-C plus aphidicolin ideal for inclusion in future clinical trials.

Comparison of P-glycoprotein expression and function with in vitro sensitivity to anthracyclines in AML.

The importance of P-glycoprotein (P-gp) in AML has been well documented. Resistance to the anthracyclines can be overcome by several agents including Cyclosporin A (CsA), PSC833 and GF120918. We describe an investigation into the expression, using MRK16 and UIC2, and function of P-gp using daunorubicin with and without modulators by flow cytometric analysis on previously frozen blast cells from 27 patients with primary or secondary AML. We compared this with the in vitro chemosensitivity, using the MTT assay, of fresh blast cells from the same patients. Whilst we found a correlation between P-gp function using CsA and GF120918 and expression using MRK16 (p < 0.05) and (p < 0.02) respectively, we were unable to find any overall correlation between expression and function of P-gp with either in vitro sensitivity to the anthracyclines, previous treatment, or 1 degree or 2 degrees disease. However it was possible to identify individual patients whose cells exhibited P-gp expression and function teamed with in vitro resistance to, and modification of, the anthracyclines. Furthermore, it is possible to identify which modulator had the greatest effect. The fact that we obtained higher indications of resistance reversal using the MTT assay along with finding P-gp expression and function in patients sensitive to the anthracyclines, suggests studies of P-gp should be teemed with chemosensitivity testing to identify specific patients who will benefit.

Evidence for the involvement of the glutathione pathway in drug resistance in AML.

We have studied altered drug detoxification through the glutathione pathway as a possible mechanism of resistance in 38 patients with AML. GST alpha, mu and pi expressions were determined using immunocytochemistry, the median percentages of positive cells being 73% (range 0-98), 55% (range 0-99) and 97% (range 80-100) respectively. MRP expression was measured using MRPm6 MoAb and flow cytometry. Results were expressed as the ratio of fluorescence associated with MRP over that of an isotype matched control (median, 1.32; range 0.95-2.15). Statistical analyses showed a significant increase in GST alpha expression in blast cells showing in vitro resistance to doxorubicin, with a median value of 78% positive cells compared to 41% in the sensitive group (p < 0.02). There was a significant reduction, however, in GST mu expression from a median value of 60% in newly presenting patients to 40% in a group of patients who had received previous cytotoxic therapy (p < 0.02). Interestingly, patients with high GST mu expression appeared to co-express MRP (p < 0.05). In vitro drug modulation studies, comparing the cytotoxic effect of doxorubicin +/- ethacrynic acid at 6.5 microM resulted in only one significant increase in sensitivity (2.6-fold), out of 22 comparisons. These results support the theory that altered detoxification through the glutathione pathway contributes towards drug resistance in AML. Further studies using fresh blast cells are required to elucidate the importance of this mechanism for individual patients.

Comparison of BCL-2 and BAX protein expression with in vitro sensitivity to ARA-C and 6TG in AML.

Activity of BCL-2 protein may be antagonised by BAX protein expression, thereby affecting cellular sensitivity to chemotherapeutic drugs. We analysed the BCL-2 protein expression of blast cells from 19 patients by flow cytometry and immunocytochemistry. This was compared to in vitro sensitivity to the anthracyclines and antimetabolites using the MTT assay. We found a significant correlation between BCL-2 expression and in vitro response to two antimetabolite drugs. One of 7 patients (14%) whose cells were sensitive to ara-C expressed BCL-2 compared to 4/4 patients (100%) whose cells were resistant to ara-C in vitro (p = 0.05). Furthermore, none of the three patients whose cells were sensitive to 6-TG expressed BCL-2 compared to 6/9 patients (67%) whose cells were resistant in vitro (p = 0.045). We found no other correlation between BCL-2 expression and any other chemotherapeutic drug analysed. The ratio of BCL-2 to BAX may be more relevant clinically, therefore cells from a further 9 patients were analysed for both proteins. Whilst there was no overall relationship between BCL-2/BAX ratios and sensitivity to ara-C and 6TG, individual patients could be identified whose blast cells were resistant to ara-C and had high BCL-2/BAX ratios. Further analysis of the significance of these ratios to drug resistance may be of future prognostic value.

Aphidicolin markedly increases the in vitro sensitivity to ara-C of blast cells from patients with AML.

Drug resistant cells often have an increased capacity to repair their DNA after damage by cytotoxic agents. Aphidicolin can inhibit this DNA repair. We describe a study of the effect of aphidicolin to modulate the sensitivity to cytotoxic drugs of blast cells from 13 patients with AML, 11 with de novo disease on presentation and 2 secondary to MDS. Three patients had relapsed following previous therapy and samples were received from 1 patient both on presentation and relapse. Blast cells were exposed to anthracyclines, antimetabolites or etoposide +/- aphidicolin (15 microM) for 48 hours. The MTT assay was used to measure cell survival and the LC50 (concentration of drug required for 50% cell kill) was calculated. Overall, there was a significant increase in sensitivity to ara-C on co-incubation with aphidicolin in 12/14 samples (p = 0.007). The median increase in sensitivity was 3.88-fold (range 1.26- to 80-fold). Interestingly, when patients were grouped according to in vitro sensitivity to ara-C, cells from resistant patients demonstrated the greatest increase in sensitivity (median 14-fold compared to 2-fold for the sensitive group, p = 0.02). Despite the documented evidence for altered DNA repair as a mechanism of resistance to the topoisomerase II inhibitors, we found no significant increase in sensitivity to daunorubicin, doxorubicin or etoposide on co-incubation with aphidicolin. Nevertheless, we believe the unparalleled modulation of ara-C warrants further investigation.

In vitro sensitivity to the liposomal preparation, DaunoXome in CLL.

In order to determine the efficacy of liposomal encapsulated daunorubicin (DaunoXome; DNX) in chronic Iymphocytic leukaemia, the sensitivities of cells from 10 patients with this disease were assessed and compared with that of free drug using the MTT assay. There was a marked variation in effect between patients for both drug preparations. Despite this, there was a small but significant enhancement of cytotoxicity afforded by the liposomal preparation (median 2.8-fold, p < 0.01). The cells from 2 of the patients appeared to be resistant to free daunorubicin. However, incubation for 96 h in DNX appeared to circumvent this resistance. This increase in sensitivity for resistant cells could not be demonstrated in an MDR positive cell line (K562AR) suggesting that another mechanism of resistance may be involved. We conclude that it is feasible to assess sensitivity to DNX using the MTT assay in fresh cells from patients with CLL using a continuous drug exposure of 96 h. Furthermore, DNX appears more effective than free drug in this disease.

Nitric oxide synthase activity in fresh cells from ovarian tumour tissue: relationship of enzyme activity with clinical parameters of patients with ovarian cancer.

Recent studies suggest a dual role for nitric oxide (NO) in tumour biology. High concentrations of NO can mediate tumouricidal activity, whereas lower concentrations have been shown to promote tumour growth. In this study, NO synthase (NOS) activity was investigated in cells that were prepared from tissue from primary and metastatic sites and from malignant effusions in 41 cases of suspected ovarian cancer. NO biosynthesis, determined by nitrite + nitrate (NOx) accumulation in medium from cultured cells prepared from disaggregated tumours or effusions and indicative of the inducible NO synthase isoform, was detected in 37% of the cases investigated (range 10.2-114 microM). There was a significant relationship between NOx and tumour differentiation (P = 0.014), with NOx being significantly higher for the more differentiated tumours. NOS activity, determined by the conversion of radiolabelled L-arginine to citrulline by tissue or cell extracts, was detected in 29% of cases (range 0.9-6.9 pmol/min per mg of protein), with all samples tested being moderately or poorly differentiated. Seventy percent of this activity was calcium dependent, indicative of constitutive NOS isoforms. Morphological and immunohistochemical assessment of tumour samples indicated a significant relationship between high macrophage content and NOS activity (as NOx biosynthesis) (rs = 0.726, N = 16, P < 0.01). The relationship between NOS expression, immune response, and disease progression is complex and not simply dependent on the differentiation status of ovarian cancer.

Modulation of resistance to ara-C by bryostatin in fresh blast cells from patients with AML.

Bryostatin has shown promise both as a cytotoxic agent and more recently as a modulator of 1-beta-D-arabinofuranosylcytosine (ara-C) resistance. This compound is currently in phase I and II trials as a single agent. We have used the 3-4,5-dimethylthiazol-2,5-diphenyltetrazolium bromide (MTT) assay as a means of investigating the direct effects of bryostatin and the effects of co-incubating this agent with ara-C on fresh blast cells from 53 patients with acute myeloid leukaemia (AML) and myelodysplastic syndrome (MDS). Additional studies evaluated the levels of accumulation and retention of 1-beta-D-arabinofuranosylcytosine 5'-triphosphate (ara-CTP) in cells exposed to ara-C with and without bryostatin. Cells were exposed to bryostatin at a range of concentrations (0.1-100 nM) for 48 h and at 1 nM for both modulation studies and assessment of ara-CTP production. We found bryostatin to be cytotoxic in 18/58 (31%) tests whilst potentiation of formazan production in the MTT assay was seen in 21/58 (36%) patients. On co-incubation with bryostatin, 16/58 (27%) tests showed increased cytotoxicity to ara-C. Furthermore, there was a significant increase in the accumulation of ara-CTP on co-incubation with bryostatin (p = 0.0401). We found patients with in vitro resistance were more likely to become sensitised following exposure to bryostatin (p < 0.01). This study has emphasised the need to optimise treatment regimens for individual patients using this approach.

The clinical relevance of chemosensitivity testing in ovarian cancer.

The aim of this study was to assess the role of in vitro chemosensitivity testing in ovarian cancer using the MTT assay. Cells were separated from solid biopsy or ascitic fluid samples from 73 patients with ovarian adenocarcinoma, FIGO stage III to IV, on presentation. A 48 h drug exposure was followed by the MTT assay to determine sensitivity. Patients were treated by conventional regimens containing platinum. There was a marked variation in sensitivity to the platinum drugs between individual patients. Clinical data were available for 37 patients. Eleven out of seventeen (65%) patients in the sensitive group had a complete response to therapy, compared with 3 out of 20 (15%) in the resistant group (p = 0.005). The overall survival rates were 36% for the sensitive group compared with 18% for the resistant group (p = 0.37). This study suggests that chemosensitivity testing in ovarian cancer may be effective in improving initial clinical response.

Aphidicolin markedly increases the platinum sensitivity of cells from primary ovarian tumours.

Enhanced DNA repair has been observed in cisplatin-resistant ovarian cancer cell lines. This resistance can be modulated, on co-incubation with aphidicolin in established cell lines and animal tumour models, by inhibiting DNA polymerases. We describe a study of the in vitro modulation effect of aphidicolin on cisplatin and carboplatin using fresh cells harvested from biopsy samples or ascitic fluids from 25 patients with ovarian adenocarcinoma. The MTT assay was used to measure cell survival after drug exposure. Aphidicolin (up to 30 microM) showed no cytotoxicity when tested alone. Forty-seven comparisons were made between drug with and without aphidicolin, and 37 (79%) cases demonstrated a significant increase in sensitivity to the platinum agents on co-incubation. Overall, there was a median 10-fold (range 1.64- to 58.5-fold) increase in sensitivity. When patients were grouped according to in vitro sensitivity to platinum, aphidicolin had a significantly greater effect in the "resistant' group, causing a median 13.5-fold increase in sensitivity compared with 2.4-fold in the "sensitive' group. Furthermore, a positive correlation between the LC50 for platinum and the corresponding fold increase in sensitivity suggests that aphidicolin overcomes platinum resistance in fresh cells from primary tumours. These results encourage the further development of this interesting compound.

The activity of N-(hydroxymethyl) melamines in fresh human ovarian tumour cells and xenografts.

Trimelamol (TM) was developed as a water soluble analogue of the oral chemotherapeutic agent hexamethyl-melamine (HMM), to be administered i.v., in an effort to avoid dose limiting emesis. Because of formulation difficulties due to its inherent instability the development of TM was halted. In vivo studies using a human ovarian cancer xenograft model PXN/65 showed TM to be curative in the dose range of 15-60 mg/kg i.p. daily x 5, for 4 weeks. Conversely, HMM given at the highest dose (60 mg/kg i.p., or 90 mg/kg p.o.) indicated only very modest tumour growth delays. In vitro chemosensitivity testing using primary ovarian tumour cultures showed that in 12/23 cases indicating reduced sensitivity to cisplatin or carboplatin, sensitivity to TM was increased. TM was curative in the carboplatin-resistant HX 110P human ovarian cancer xenograft and promising activity was seen in the MX-1 human breast cancer xenograft. In spite of enhanced stability in aqueous solution and good in vitro cytotoxicity, the TM analogues CB 7669 (triscyanomethyl) and CB 7639 (tristrifluoroethyl) showed disappointing in vivo antitumour activity which may be explained by the need for prolonged exposure. TM analogues with intermediate stability are currently under development in an effort to further the clinical development of this promising group of antitumour agents.

An in vitro study of blast cell metabolism in acute myeloid leukaemia using the MTT assay.

We have utilized the MTT assay to measure the metabolic activity of cells from the bone marrow of 55 patients with acute myeloid leukaemia (AML), myelodysplastic syndrome (MDS) and non-clonal disease. Doubling dilutions of cells were exposed to MTT for 3-4 h. The mean optical density of the formazan produced by each cell dilution was plotted and the gradient of the line produced was calculated, higher gradients indicating more metabolically active cells. Results showed that the median activity of mononuclear cells from seven patients with non-clonal disease was 0.202 (range 0.175-0.253); blast cells from 27 patients with de novo AML had a median activity of 0.187 (range 0.079-0.345) and 13 patients with MDS a median of 0.155 (range 0.062-0.311). Seven assays on mononuclear cells from five patients in remission had a median activity of 0.203 (range 0.190-0.248), indicating no significant difference between these and normal patients. There was no correlation between the metabolic activity of cells when compared with their proliferative capacity, cell size and expression of P-glycoprotein. Following exposure of the AML patients' blast cells to the anthracyclines, cytosine arabinoside, 6-thioguanine and etoposide, cell survival was measured using the MTT assay. While there was no correlation between the in vitro sensitivity of these cells to the anthracyclines or etoposide, less metabolically active cells showed significantly greater sensitivity to 6-thioguanine. Conversely, the more active cells appeared to be more sensitive to cytosine arabinoside. Patients whose blasts cells showed higher metabolic activity appeared to achieve remission and had a longer mean survival time. Therefore, by using a simple technique we were able to establish that some patients were more likely to respond to certain cytotoxic regimes. Our preliminary study reflected the multifactorial nature of clinical response in AML and MDS, so providing further information on the relationship between cellular metabolic activity and treatment failure.

Comparison of P-glycoprotein expression with in vitro drug sensitivity in fresh blast cells from acute myeloid leukaemia patients.

Anthracyclines and etoposide have been implicated in the multi-drug resistance phenotype. The mdr 1 gene encodes for the transmembrane protein P-glycoprotein. P-glycoprotein expression was measured in the fresh blast cells from 19 patients with acute myeloid leukemia using three monoclonal antibodies, C219, JSB-1 and MRK 16, and immunocytochemistry with the enzyme alkaline phosphatase as marker. Drug resistance can be identified in vitro using the predictive chemosensitivity test, the MTT (3-4,5-dimethylthiazol-2,5-diphenyl tetrazolium bromide) assay. In order to assess cell viability after drug exposure, this technique utilises the ability of cellular dehydrogenase enzymes to reduce the tetrazolium salt MTT to formazan. In vitro resistance to multi-drug resistance related cytotoxic agents was identified in the blast cells from these patients. This study showed no correlation between the results of the MTT assay and P-glycoprotein expression in this disease, suggesting either that more sensitive techniques are required to measure P-glycoprotein expression or that other drug resistance mechanisms may be involved.

A feasibility study of the MTT assay for chemosensitivity testing in ovarian malignancy.

We assess the feasibility of using the MTT assay as a measure of cell viability in chemosensitivity testing in ovarian malignancy. The assay utilises the conversion of the tetrazolium salt MTT to formazan by dehydrogenase enzymes in living cells. We show that the optical density of the formazan produced from MTT is directly proportional to the number of live cells tested. Optimum MTT conversion occurred after 4 h incubation and dimethyl sulphoxide was found to be the most suitable solvent for the formazan. Seventy-five samples of ascitic fluid and/or solid tumour were collected from 56 patients with FIGO stage III-IV ovarian adenocarcinoma. Malignant cell suspensions with a viability greater than 75% were prepared from 95% of ascitic fluid and 75% of biopsy samples by simple techniques. The effect of cytotoxic drugs was assessed in 91% of patients included in the study. Variation in drug effect between patients was evident following a 48 h incubation period and was reproducible. Overall platinum and anthraquinone analogues produced the greater effect but resistance did occur. Our results mirrored reported clinical response rates. Only one sample tested against chlorambucil showed any drug effect. As this assay produces results in a high percentage of tests and is rapid and simple it appears suitable for prospective clinical trials to correlate the in vitro results with in vivo response.