Tumor excisions or biopsies were processed within 24 to 48 hours of collection. The specimen was minced with scalpels and tissue pieces were digested with 0.25% trypsin and 0.08% DNase for one to two hours at 37℃. Non-digested tissue was removed by filtration through a 100-micrometer cell strainer. When necessary, non-viable cells were removed by density gradient centrifugation. The resulting viable cell suspension was then incubated for 30 minutes at 37℃ in a tissue culture flask to remove macrophages by adherence. Non-adherent cells were collected and the remaining hematopoietic cells were depleted after 30 minutes incubation with specific antibody-coated magnetic beads. The final cell suspension was plated into a 96 well half-area plate, 120 microliter aliquot per well. The plate was incubated overnight at 37℃ with 5% carbon dioxide humidified atmosphere. The cell seeding concentration varied between 5×10⁴ to 1.5×10⁵ cells per well. It was adjusted depending on cell volume to give adequate well coverage.

The purity of the cell preparation was analyzed by immunohistochemistry and Wright-Giemsa staining to confirm malignant cell content and phenotype. A pathologist evaluated each specimen to be certain all tumors were endometrial carcinoma. To be deemed acceptable for the MiCK assay, the final cell suspension should contain at least 80% of tumor cells. Based on our experience, 20% or less nonmalignant cells in the suspension is not enough to interfere with the results.3

Human JURL-MK2 chronic leukemia blast crisis cell line (DSMZ, Braunschweig, Germany) was used as a positive control for MiCK assays done with patient tumor cells. RPMI-1640 medium without phenol red was used for all cultures. It was supplemented with 10% fetal bovine serum, 100 units/mL penicillin, and 100 micrograms/mL streptomycin. Cell counts and viability were evaluated by trypan blue exclusion dye with a hemocytometer.

The MiCK assay procedure was adapted from the method described previously.1,3,6 The purified tumor cells were incubated overnight at 37℃ in a 5% carbon dioxide humidified atmosphere incubator to allow the cells to adapt to the culture environment and attachment to the culture plate. Cell growth is not necessary for the MiCK assay. After this incubation, the chemotherapy drugs were added in the wells of the 96 well plate in 5 microliter aliquots at various concentrations. Single and combination regimens were tested; including the combination of doxorubicin+cisplatin+paclitaxel and the combination of carboplatin+paclitaxel (arms for the current GOG endometrial cancer phase III trial GOG 209) as well as single agent testing with paclitaxel, carboplatin, doxorubicin, ifosfamide (4-hydroperoxyifosfamide was used as a surrogate active metabolite) and vincristine (all active agents in GOG trials). Following drug addition, the plate was incubated for 30 min at 37℃ in a 5% carbon dioxide humidified atmosphere incubator. Each well was then overlaid with mineral oil, and the plate was placed into the incubator chamber of a microplate reader (BioTek instruments, Winooski, VT, USA). The optical density at 600 nanometers was monitored every 5 minutes over a period of 48 hours. The cells remained in contact with the drugs for the entire 48 hours of the test.

All data acquisitions and calculations during the MiCK assay were done with DiaTech Oncology proprietary software ProApo. To provide a kinetic representation of the drug responses, the optical density (OD) readings were plotted against time. Apoptosis is characterized by a steep increase in OD while necrosis shows a decline in OD values.2 The MiCK assay can therefore distinguish apoptotic versus necrotic cell death. For convenience, the measure of apoptosis is reported in kinetic units (KU) as described in detail previously.3 Briefly, the maximum slope of the apoptotic curve (Vmax) is calculated for each plot of drug treated microculture. It is then compared to the Vmax of a control culture without drug. For convenience the Vmax is multiplied by 60 to convert the units from mOD/minute to mOD/hour. Finally, the data are normalized by multiplying with a standard cell density coefficient value (SCDC). The SCDC is calculated as follows: SCDC = 0.09/(OD_ctrl-OD_blank). The 0.09 was determined experimentally as the optimal OD difference between the initial OD given by the tumor cells and a blank well containing only the cell culture medium. The KU are therefore calculated with this formula:

KU: (Vmax_drug-treated - Vmax_ctrl) × 60 × 0.09/ (Mean OD_ctrl - Mean OD_blank).

In initial evaluations, the chemosensitivity response was considered high when the value was above 4.0 KU, intermediate between 1.0-4.0 NKU, and not sensitive when below 1.0 KU.

MiCK KU data for different drugs and different doses was logged into Microsoft Excel. Data were imported into SAS/JMP for analysis.10 Concentrations lower than the clinical dose were routinely used in the assay as an internal control where less apoptosis would be induced by lower concentrations. The highest value of apoptosis was given by one of the two highest concentrations that were closest to the clinical doses. SAS/JMP was used to calculate summary statistics and perform statistical analysis. Subtracting one drug's KU value from another drug's KU value on a matched sample basis made comparisons between drugs. The nonparametric Wilcoxon test was then used to test for a statistical difference.11