A general native-state method for determination of proliferation capacity of human normal and tumor tissues in vitro

Diagnosis and Pathological Analysis of Patient Cancers by Detection of Proliferating Cells in Gelfoam® Histoculture

Patient tumors grew in Gelfoam^® histoculture with maintenance of tissue architecture, tumor-stromal interaction, and differentiated functions. In this chapter, we review the use of Gelfoam^® histoculture to demonstrate proliferation indices of major solid cancer types explanted directly from surgery. Cell proliferation was visualized by histological autoradiography within the cultured tissues after [³H]thymidine incorporation by the proliferating cells. Epilumination polarization microscopy enables high-resolution imaging of the autoradiography of each cell. The histological status of the cultured tissues can be assessed simultaneously with the proliferation status. Carcinomas were observed to have areas of high epithelial proliferation with quiescent stromal cells. Sarcomas have high proliferation of the cancer cells of mesenchymal organ. Normal tissues can also proliferate at high rates. Mean growth fraction index (GFI) was highest for patient tumors with the pure subtype of small-cell lung cancer than other types of lung cancer.

Expression and Targeting of Tumor Markers in Gelfoam® Histoculture: Potential Individualized Assays for Immuno-Oncology

Tumor-specific antigens are important in the study of tumor biology, tumor diagnosis, and prognosis and as targets for tumor therapy. This chapter reviews patient colon, breast, and ovarian tumors in 3-dimensional Gelfoam^® histoculture maintaining in vivo-like expression of the important tumor antigens, for example TAG-72 and CEA. We have also reviewed that fluorescent antibodies can target tumors in Gelfoam^® histoculture, thereby providing an assay for individual patients for sensitivity to therapeutic antibodies which have become so important in immuno-oncology and other cancer therapies.

Novel ex vivo ovarian cancer tissue explant assay for prediction of chemosensitivity and response to novel therapeutics

The majority of ovarian cancer patients present with advanced disease and despite aggressive treatment, prognosis remains poor. Response to first-line carboplatin-containing chemotherapy is usually good, however, recurrence rates and subsequent chemoresistance are very high and ultimately responsible for the fatal outcome of the disease. To improve treatment outcomes pre-clinical models that can predict individual patient response to 1st line chemotherapy and novel therapeutics are urgently required. In this study, we employed an ex vivo ovarian cancer tissue explant assay to assess response to carboplatin and an inhibitor of the extracellular matrix molecule, hyaluronan (4-methylubelliferone, 4-MU), shown to inhibit cancer metastasis. Cryopreserved ovarian cancer tissues were cultured on gelatine sponges for 48-120 h with increasing concentrations of carboplatin (0-400 μM) or 4-MU (1 mM) alone or the combination of both drugs. Effects on apoptosis and proliferation were assessed by immunohistochemistry using antibodies to cleaved caspase 3 or Ki67, respectively. The ex vivo tissue explant assay maintained viable tumor cells in an intact tumor microenvironment similar to the in vivo situation over the 120 h culture period. Carboplatin treatment promoted apoptosis in chemosensitive (P = 0.0047) but not chemoresistant cancer tissues. The combination of 4-MU (1 mM) and carboplatin (100 μM) significantly increased apoptosis (P = 0.0111) and reduced proliferation (P = 0.0064) in chemoresistant tissues. Overall, our results show that the ex vivo explant assay is a robust and cost effective model to assess chemosensitivity and the effect of novel therapeutics in ovarian cancer.

Clinical Correlation of the Histoculture Drug Response Assay for Head and Neck Cancer

Gelfoam^® histoculture was utilized to develop the histoculture drug response assay (HDRA) for head and neck cancer. Specimens of head and neck tumors were evaluated for sensitivity to the following drugs: cisplatinum (CDDP), 5-fluorouracil (5-FU), and the combination of CDDP and 5-FU. In the first clinical study at UCSD, 10 of 12 patients with tumors that were drug sensitive in Gelfoam^® histoculture had either complete or partial response clinically. Comparisons of HDRA results, obtained with [³H]thymidine incorporation as the endpoint were made with clinical responses, i.e., complete response, partial response, or no response. The overall accuracy of the HDRA was 74% in this correlative clinical trial; the predictive positive value was 83%, the sensitivity was 71%, and the specificity was 78%. Seven of 11 patients with HDRA-resistant tumors demonstrated no response for a predictive negative value of 64%. In a subsequent study at Memorial Sloan Kettering Cancer Center, tumor specimens from 41 to 42 patients undergoing treatment for head and neck cancer were successfully evaluated by the HDRA. The histocultured tumors were treated with 5-FU and/or CDDP and a control group received no drug treatment. After completion of drug treatment, the relative cell survival in the tumors was determined using the MTT endpoint. Sensitivity was defined as a tumor inhibition rate (IR) of greater than 30%. Survival comparisons were performed using the generalized Wilcoxon test for the comparison of Kaplan-Meier survival curves. Resistance to 5-FU was observed in 13 cases (32%), to CDDP in 13 cases (32%), and to both agents in 11 cases (27%). The 2-year cause-specific survival was significantly greater for patients sensitive to 5-FU than patients who were resistant (85% vs. 64%), CDDP (86% vs. 64%), or both agents (85% vs. 63%). These results demonstrate the clinical usefulness of the HDRA for head and neck cancer.

A comparison of vitamin D activity in paired non-malignant and malignant human breast tissues

Links between a low vitamin D status and an increased risk of breast cancer have been observed in epidemiological studies. These links have been investigated in human tissue homogenates and cultured cell lines. We have used non-malignant, malignant and normal reduction mammoplasty breast tissues to investigate the biological and metabolic consequences of the application of vitamin D to intact ex vivo human breast tissue. Tissues were exposed to 1α,25(OH)(2)D(3) (1,25D; active metabolite) and 25(OH)D (25D; pre-metabolite). Changes in mRNA expression and protein expression after vitamin D exposure were analysed. Results indicate that while responses in normal and non-malignant breast tissues are similar between individuals, different tumour tissues are highly variable with regards to their gene expression and biological response. Collectively, malignant breast tissue responds well to active 1,25D, but not to the inactive pre-metabolite 25D. This may have consequences for the recommendation of vitamin D supplementation in breast cancer patients.

Combination of p53 codon 72 polymorphism and inactive p53 mutation predicts chemosensitivity to 5-fluorouracil in colorectal cancer

There are increasing reports showing the clinical significance of the p53 polymorphism status in terms of the response to chemotherapy. We investigated whether p53 polymorphism and mutation were associated with in vitro sensitivity to 5-fluorouracil (5-FU) in patients with colorectal cancer. Chemosensitivity to 5-FU was evaluated by the collagen gel droplet embedded culture drug sensitivity test. 5-FU sensitivity of tumor cells without inactive p53 mutation in the arginine/arginine (Arg/Arg) variant was significantly higher than that of tumor cells with or without inactive p53 mutation in other variants (p = 0.022), whereas the 5-FU sensitivity of tumor cells with inactive p53 mutation in the Arg/Arg variant was significantly lower than that of tumor cells with or without inactive p53 mutation in other variants (p = 0.002). In the Arg/Arg variant, apoptotic cells induced by 5-FU treatment in patients without inactive p53 mutation were more markedly increased than those in patients with inactive p53 mutation (p = 0.037). Bax and Bcl-2 protein expressions in tumor tissue treated with 5-FU were associated with both 5-FU sensitivity and the apoptotic cell count. Our data show that the Arg/Arg genotype without inactive p53 mutation could be predictive of a more favorable response and the Arg/Arg genotype with inactive p53 mutation a less favorable response to chemotherapy using 5-FU in CRC. The combination of the p53 codon 72 polymorphism and p53 mutation status is a potential predictive marker of sensitivity to 5-FU in CRC.

Evaluation of 5-fluorouracil applicability by the collagen gel droplet embedded drug sensitivity test with area under the curve analysis

We have evaluated the 5-fluorouracil sensitivity of cancer cells from colorectal cancer patients using the collagen gel droplet embedded drug sensitivity test under multiple drug concentrations and contact durations. After converting drug concentration and contact time to the area under the curve (AUC) and plotting against the growth inhibition rate, the correlation between AUC and the growth inhibition rates was approximated to the logarithmic regression curve. In this study, to further validate the reliability of the regression curve, the growth inhibition rate was calculated from the regression curve and the actual growth inhibition rate was compared at AUC of 48 mug h/ml. No significant difference was observed in the growth inhibition rates between the two groups by paired t-test (P=0.590). A strong positive correlation was found between the two groups by regression analysis (y=0.7555x+10.514, R=0.8236). This result strongly suggests that in-vitro antitumor effect of 5-fluorouracil depends on the AUC in colorectal cancer and the AUC-inhibition rate curve is reliable. We can obtain the inhibition rate from AUC and vice versa using the AUC-inhibition rate curve. We can also calculate the individualized AUCIR50, AUC value that gives 50% growth inhibition, using the AUC-inhibition rate curve. This could be useful to establish individualized chemotherapy using the collagen gel droplet embedded drug sensitivity test.

Sensitivity to cisplatin determined by the histoculture drug response assay and clinical response of endometrial cancer

This study investigated the value of the in vitro histoculture drug response assay (HDRA) for predicting the efficacy of chemotherapy in patients with endometrial cancer. Specimens were obtained from 115 patients with endometrial cancer treated at Keio University Hospital between 1994 and 2002. Tumor fragments were cultured on collagen sponge gel with cisplatin for 7 days, and cell viability was assessed. The cutoff value of the 50% inhibitory concentration of cisplatin was set at 23 microg/mL. Sensitivity of stage III or IV disease to chemotherapy was investigated, and differences of 5-year progression-free survival between patients with sensitive and resistant tumors were evaluated by the Kaplan-Meier method. Tumors were evaluable in 93.0% of patients (107/115). Among 38 patients in stages III or IV, 23 received chemotherapy containing cisplatin. Seven sensitive tumors did not recur, while recurrence/progression occurred within 6 months in 8/16 patients with tumors showing low sensitivity. Among stages III and IV patients, there was a significant difference of 5-year progression-free survival (P < 0.05) between those with tumors showing high or low sensitivity. Accordingly, the HDRA may predict the efficacy of chemotherapy for endometrial cancer.

Examination of in vitro chemosensitivity test using collagen gel droplet culture method with colorimetric endpoint quantification

To develop a simpler method of performing the collagen gel droplet-embedded culture drug sensitivity test (CD-DST), we examined the introduction of colorimetric quantitative determination of images for evaluation of anticancer effect against cancer cells alone in the presence of fibroblasts, based on differences in proliferative morphology and stainability with neutral red of cells within collagen gel drops determined using a video-microscope and NIH Image software. In examinations using a human cancer cell line and a fibroblast cell line, a high degree of linearity between number of cancer cells and image-optical density was found within the range of 10(2) - 10( 6) cells / droplet (r (2) = 0.933). Using NIH Image, fibroblast cells could be eliminated at a cut-off value of 128, and an immunocytochemical method demonstrated that the cells eliminated from the image were indeed fibroblasts, and those remaining were cancer cells. CD-DST was carried out with mixtures of cancer cells with fibroblasts at various ratios, and the feasibility of evaluating anticancer activity in cancer cells alone with no effect of fibroblasts at any mixing ratio was confirmed. In addition, for CD-DST of primary cell cultures of human lung cancers collected at the time of surgery, a high correlation between results obtained with the volume supplementation method, a current cell quantification method, and those with the imaging colorimetric quantification method was obtained (r = 0.933). These results indicate that introduction of imaging colorimetric quantification utilizing NIH Image makes CD-DST a quick and simple method that should be highly useful for clinical chemosensitivity testing using primary cell cultures of human cancers.

New sensitive discovery histoculture model for growth-inhibition studies in prostate cancer and BPH

BACKGROUND

A new, total-immersion three-dimensional histoculture (TIH) method was developed to evaluate growth of tissue containing a mixture of benign prostate hyperplasia (BPH) and prostate cancer in vitro.

METHODS

Efficacy of inhibitors, such as genistein, was determined by measuring 3H-thymidine incorporation per microgram protein. Inhibitory effects obtained in TIH were compared to those in sponge-gel supported histoculture (SSH).

RESULTS

3H-thymidine incorporation was 2-5-fold higher in tissue cultured in TIH than in SSH. The average inhibition by genistein at a concentration of 18 JIM was 73% in TIH, vs. 31% in SSH. TIH also appeared to be more sensitive, since the lowest concentration of genistein that significantly inhibited growth of BPH mixed with prostate cancer tissue was 2.3 IJM, while in SSH the lowest concentration was 9.2 F,M. Although the within-assay coefficient of variation (CV) was similar for both TIH and SSH, the between-assay CV was better in TIH.

CONCLUSIONS

These data suggest that TIH can be used as a discovery model for screening and evaluating inhibitors of prostate tissue growth in vitro.

Governing step of metastasis visualized in vitro

Metastasis is the ultimate life-threatening stage of cancer. The lack of accurate model systems thwarted studies of the metastatic cell's basic biology. To follow continuously the succeeding stages of metastatic colony growth, we heritably labeled cells from the human lung adenocarcinoma cell line ANIP 973 with green fluorescent protein (GFP) by transfection with GFP cDNA. Labeled cells were then injected intravenously into nude mice, where, by 7 days, they formed brilliantly fluorescing metastatic colonies on mouse lung [Chishima, T., Miyagi, Y., Wang, X., Yang, M., Tan, Y., Shimada, H., Moossa, A. R. & Hoffman, R. M. (1997) Clin. Exp. Metastasis 15, 547-552]. The seeded lung tissue was then excised and incubated in the three-dimensional sponge-gel-matrix-supported histoculture that maintained the critical features of progressive in vivo tumor colonization while allowing continuous access for measurement and manipulation. Tumor progression was continuously visualized by GFP fluorescence in the same individual cultures over a 52-day period, during which the tumors spread throughout the lung. Histoculture tumor colonization was selective for lung cancer cells to grow on lung tissue, because no growth occurred on histocultured mouse liver tissue, which was also observed in vivo. The ability to support selective organ colonization in histoculture and visualize tumor progression by GFP fluorescence allows the in vitro study of the governing processes of metastasis [Kuo, T.-H., Kubota, T., Watanbe, M., Furukawa, T., Teramoto, T., Ishibiki, K., Kitajima, M., Moossa, A. R., Penman, S. & Hoffman, R. M. (1995) Proc. Natl. Acad. Sci. USA 92, 12085-12089]. The results presented here provide significant, new opportunities to understand and to develop treatments that prevent and possibly reverse metastasis.

"Decadose" effects of cisplatin on squamous cell carcinoma of the upper aerodigestive tract. I. Histoculture experiments

There is substantial laboratory and clinical evidence that solid tumors rapidly acquire cellular resistance to cisplatin. Experiments with human carcinoma cell lines and clonogenic assays indicate that resistance is usually mild to moderate and can be circumvented with higher concentrations of drug. The purpose of this investigation was to test this hypothesis with a histoculture assay of human upper aerodigestive tract (UADT) carcinomas. Using a sponge-gel supported histoculture, 43 tumor specimens from patients with squamous cell carcinoma (SCC) of the UADT were grown and exposed to cisplatin. Growth inhibition by the drug, in concentrations equivalent to peak therapeutic doses (1.5 micrograms/mL) and concentrations 10 and 25 times greater (15 and 37.5 micrograms/mL), were measured in specimens from patients with previously untreated and recurrent lesions. In vitro, the overall rate of sensitivity of the tumor samples to cisplatin concentrations of 1.5, 15, and 37.5 micrograms/mL were 22%, 62%, and 83%, respectively. In patients with previously untreated disease, the respective rates were 25.9%, 63.3%, and 79.3%, as compared with 10.0%, 55.6%, and 85.6%, respectively, for patients with recurrent disease. The response difference between cisplatin concentrations of 1.5 and 15 micrograms/mL was statistically significant. The "decadose" effect of cisplatin on growth inhibition was 2.44-fold for untreated lesions and 5.56-fold for recurrent tumors. The results indicate that resistance to standard doses of cisplatin by SCC of the UADT can be substantially overcome with a decadose (standard dose x 10) increase and is more pronounced in tumors from patients with recurrent disease. Progress toward improving survival of patients may be possible by incorporating decadose cisplatin therapy into a multimodality treatment plan.

In vitro chemosensitivity test of human brain tumors using a three-dimensional organ culture with a collagen gel matrix

We evaluated the usefulness of an in vitro tumor organ culture system using a specialized collagen gel matrix derived from pig skin as a chemosensitivity test for human brain tumors. Two xenograft tumors derived from human glioblastoma and medulloblastoma were examined with this system and the results were compared with data obtained from a nude mouse assay. Xenograft tumors exhibited in vivo-like three-dimensional growth on the collagen gel matrix and had increasing incorporation of tritiated (3H)thymidine for 2 weeks. Drug sensitivity, as measured by this assay at therapeutic peak plasma concentrations of anticancer drugs, corresponded with that measured with the nude mouse assay. Chemosensitivity of 16 surgical specimens of malignant brain tumors were also examined successfully by this collagen gel matrix (CGM) assay. When the highest inhibition rate in dose-inhibition curve was equal to or greater than 50%, the tumor was regarded to be sensitive to the agent. The efficacy rates in CGM assay for 16 lesions were 25.0% (4/16) for ACNU, 67.8% (11/16) for adriamycin, 31.3% (5/16) for cisplatin, and 67.8% (11/16) for etoposide. The CGM assay has advantages as a chemosensitivity test because of its simple procedure, rapidity, high rate of evaluable tumor growth, and in vivo-like three dimensional tumor growth. Our results indicate that the CGM assay is feasible to test the chemosensitivity of malignant brain tumors.

Histocultures of patient head and neck tumors for pharmacodynamics studies

This investigation was to establish a clinically relevant experimental model to evaluate the pharmacodynamics of drugs used for head and neck cancers. A total of 83 surgical samples of primary and lymph nodal metastatic tumors was obtained from 66 patients. Fragments of these tumors were cultured on a collagen gel matrix. The tumor cell labeling index (LI) was determined by [3H]thymidine incorporation and autoradiography. Seventeen tumors (20%) were contaminated. About 80% of the remaining 65 tumors were successfully cultured for at least 2 weeks. The cultured tumor fragments retained the morphology and architecture of the freshly removed specimens; both tumor and stromal cells were present. The tumor cell LI after 2-3 weeks in culture, determined from the most proliferative area of the tissue, averaged 77 +/- 12% for primary tumors and 78 +/- 12% for nodal metastases. The activity of three clinically active agents, 5-fluorouracil (FU), cisplatin (DDP), and mitomycin C (MMC), was evaluated in 47 tumors. All three drugs inhibited the tumor LI. The concentrations needed to produce a 50% inhibition of the tumor LI (IC50) were within the clinically achievable concentration range. The intertumor variation in the IC50 for FU (60-fold) was considerably greater than that for DDP and MMC (7- to 8-fold). The nodal metastatic tumors appeared to be less sensitive to FU than the primary tumors, while there were no apparent differences for DDP or MMC.(ABSTRACT TRUNCATED AT 250 WORDS)

To do tissue culture in two or three dimensions? That is the question

Alexis Carrel introduced the in vitro culture of tissues in the beginning of the century utilizing a culture system that allowed the three-dimensional growth of tissues. Leighton improved upon this system by developing a substrate of sponge matrices. Other methods of three-dimensional culture include collagen gels and what are known as organ culture systems on filters or meshes. In addition, cell suspensions can be converted into multicellular spheroids, another form of three-dimensional culture. Comparison of the three-dimensional culture methods with two-dimensional culture methods has shown critical differences in the behavior of biological systems in culture. For example, in vivo-like drug responses are observed in three-dimensional but frequently not in two-dimensional cultures, indicating that drug response may be a function of tissue architecture. The in vivo mechanism of drug resistance may involve alterations in cell-cell interaction which may occur in three-dimensional culture as opposed to monolayer culture. Practical applications of three-dimensional culture include the development of a drug-response assay that correlates not only with drug resistance but also with drug sensitivity and survival of cancer patients. It has been shown that gene expression may be more in vivo-like in three-dimensional cultures than in two-dimensional monolayer cultures. For example, tumor antigens may be expressed in three-dimensional culture and not in monolayer culture. Thus, future studies utilizing three-dimensional cultures may significantly enhance our understanding of gene expression and resistance to drugs and enhance the efficacy of cancer chemotherapy by correctly predicting active drug regimens for individual patients.

Comparative study of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide and tritiated thymidine in a chemosensitivity test using collagen gel matrix

An organ culture system using collagen gel matrix (CGM) was modified and established successfully as a rapid and convenient method for determination of anti-cancer drug sensitivities. The cell viability of a tumor fragment was measured by the reduction of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) to a colored formazan product, which allowed for quantitative and simple analysis. The sensitivities of KK-47 bladder tumor from nude mice to various anti-cancer drugs tested corresponded closely to those determined in the original CGM assay system using tritiated thymidine, which has a high clinical correlation. Our modified method can be used as a practical and highly reproducible chemosensitivity test in vitro.

Analysis of hepatocellular proliferation: study of archival liver tissue is facilitated by an endogenous marker of DNA replication

Assessment of liver regeneration with endogenous genes that are expressed during DNA replication is physiological, specific and direct. To determine whether H3 histone messenger RNA expression (which is tightly coupled with DNA synthesis) could be used for this purpose, we initially examined liver regeneration in a mouse model. After partial hepatectomy, RNA transblot studies showed induction of H3 histone messenger RNA expression in regenerating mouse livers. In situ molecular hybridization demonstrated that the overall pattern of H3 histone messenger RNA expression correlated with [3H]thymidine labeling of hepatocytes. After partial hepatectomy, H3 histone messenger RNA expression in hepatocytes peaked at 48 hr (greater than 60 times greater than at 24 hr; p less than 0.001) and then rapidly declined. Although hepatocyte labeling with [3H]thymidine showed similar kinetics of liver regeneration, use of this parameter resulted in overestimation of the proliferative compartment when it was compared with H3 histone messenger RNA expression. Next we determined whether H3 histone messenger RNA expression could be used to study hepatocellular proliferation in archival human material. H3 histone messenger RNA-expressing hepatocytes were identified on in situ hybridization in patients with acute or chronic active hepatitis and active cirrhosis, but not inactive cirrhosis. These studies demonstrate that H3 histone messenger RNA is expressed in a phasic manner during liver regeneration. Use of H3 histone messenger RNA expression to evaluate hepatocellular proliferation should facilitate clinical studies and greatly advance our understanding of the pathophysiology of liver regeneration.

Growth of human renal cortical tissue on collagen gel

A model system for 3-dimensional "native-state" culture of tissues on collagen gels (Proc. Natl. Acad. Sci. USA 86:2013-2017; 1989) has been applied in this study to histologically normal human renal cortical tissue from 11 patients undergoing nephrectomy for renal cell carcinoma elsewhere in the kidney. Microbial contamination occurred in 12/90 cultures, the rest (78) were studied by visual inspection, histology, immunohistochemical analysis for pankeratin (epithelial cell origin), vimentin (mesenchymal cell origin), and p-glycoprotein (associated with proximal tubules), transmission electron microscopy (EM), incorporation of tritiated thymidine (3HTdR). In the first 10 days, explants showed 3HTdR-labeled cells in tubule structures. The surrounding gel was invaded by cells forming tubule structures, sometimes with basement membrane. Some of these cells showed labeling by 3HTdR and immunostaining positive for pankeratin and p-glycoprotein. EM showed well-polarized epithelial cells in tubule structures with tight junctions, interdigitating lateral processes, and microvilli characteristic of proximal and distal convoluted tubules. 3HTdR-labeled cells in tubule structures were observed even 2 mo. after Passage 1, 6 mo. after the initial explantation. Tubule growth was most active and fibroblast proliferation was negligible from 2 to 4 wk postexplantation. The proliferation of tubulelike cells and formation of tubulelike structures in this system represents an opportunity to study human renal cortical tissue in vitro, under conditions more closely resembling in vivo circumstances than are present in other in vitro systems suitable for long-term study. This model has potential use for in vitro toxicology studies and studies of renal physiology.

Correlation of histology and drug response of human tumors grown in native-state three-dimensional histoculture and in nude mice

An in vitro histoculture system in which a native-state collagen-sponge gel supports the three-dimensional growth of tumor tissue has been recently developed that allows the culture and drug response assay for most every tumor type. Important features of the histoculture system include the maintenance of three-dimensional tissue architecture and the use of histological autoradiography to measure nuclear incorporation of [3H]thymidine as an endpoint. We describe in this report in vitro-in vivo correlations for drug response and tumor histology by using human tumor xenografts grown in the native-state three-dimensional histoculture system and as xenografts in nude mice. This comparison eliminates many of the confounding variables seen in most correlative clinical trials. Results demonstrate (i) a very high preservation of in vivo tissue architecture in vitro, (ii) an 86% accuracy in vitro of predicting drug resistance in vivo using suprapharmacologic doses of drugs in vitro, and (iii) an overall predictive frequency of drug resistance and sensitivity ranging from 53% for 5-fluorouracil to 78% for doxorubicin.

Organoid reorganization of human tumors under in vitro conditions

In the present study we describe a new method to cultivate human tumors, which allows organoid differentiation under in vitro conditions. Diverse tumors of different origin and various histopathology which had been heterotransplanted to athymic mice were dissociated into single cells and seeded at high cell density onto a membrane filter consisting of cellulose nitrate at the gas-medium interface. Within a few days, the tumor cells reorganized and differentiated into organoid structures which exhibited the typical histological characteristics of the original tissues. Due to the formation of organoid aggregates, which was also previously seen with normal fetal cells, this type of culture has been described as 'organoid culture'. In the case of adenocarcinomas of the lung and the colon including the rectum, glandular structures with central lumina, adjacent microvilli, and junctional complexes were formed. Numerous specific intercellular contacts such as desmosomes and tight junctions occurred as well as interdigitations of adjacent cell membranes. In a tumor of the rectum, a typical brush border differentiated at the surface of the reorganized tumor-tissue aggregate. Epidermoid carcinomas of the head and neck developed structures resembling the spinous layer of the epidermis, exhibiting numerous desmosomes and intracytoplasmic bundles of tonofilaments radiating into the desmosomes. Most tumors produced a fragmentary monolayered or multilayered basal lamina of similar morphological appearance as under in vivo conditions. These results illustrate the organoid reorganization and differentiation of human tumor cells under the experimentally rather simple conditions of the organoid culture systems and clearly demonstrate that this in vitro system comes close to the in vivo situation as far as certain differentiation phenomena are concerned.

Skin toxicity determined in vitro by three-dimensional, native-state histoculture

We describe a gel-supported in vitro system for culturing skin samples in a three-dimensional native state. All the cell types of skin remain viable and maintain their native architecture for at least 10 days. The culture system is used for toxicity measurements by ascertaining cell viability using two fluorescent dyes: 2',7'-bis-(2-carboxyethyl)-5-(and -6)carboxyfluorescein acetoxymethyl ester, specific for living cells, and propidium iodide, specific for dead cells. Cell staining with the dyes is measured throughout the tissue block by confocal scanning fluorescence microscopy. The dose-response to three agents--ethanol, doxorubicin, and sodium hypochlorite--is shown and, in the case of sodium hypochlorite, compared to in vivo skin toxicity with a high correlation. We also demonstrate that the end point of [3H]thymidine incorporation measured by histological autoradiography can be used to measure toxicity. Our results with the [3H]thymidine end point demonstrate that the hair follicle cells are the most sensitive to doxorubicin. The native-state model for skin may be an effective replacement for animal systems and superior to the dispersed skin cell systems used previously. It can allow rapid, inexpensive measurements of the effect of manufactured products, drugs, and pollutants on skin.

In vitro sensitivity assays in cancer: a review, analysis, and prognosis

Tumors are complex systems consisting of heterogeneous cancer cells as well as normal cells with each exhibiting unique drug sensitivity spectra. There have been many attempts to design in vitro systems to determine drug response to tumors. The most widely used system is the clonogenic assay. which has demonstrated some clinical predictivity. However, the clonogenic assay has been shown to have negative aspects, including low frequency of evaluation, clump artifacts, lack of cytotoxic end-points and disruption of normal cell-cell interactions existing in a true tissue environment. Newer models are described utilizing cytotoxic as well as cell-proliferation end-points, and maintenance of three-dimensional tissue architecture in vitro. It is concluded that less artifactual, more realistic models can be used to select more tumor-specific drugs which themselves in turn will make in vitro chemosensitivity assays more useful for cancer patients.

A rapid method to determine proliferation patterns of normal and malignant tissues by H3 mRNA in situ hybridization

A general method applicable for the determination of any mammalian tissue's proliferative pattern is described. This method determines the cellular mRNA level of a proliferation-dependent gene, histone H3, by in situ hybridization. The cell-cycle S-phase-specific expression of this highly conserved ubiquitous cellular gene, and the lack of it in resting cells, permits the unambiguous identification of cycling cells in any tissues, normal or diseased. This method can be conveniently coupled with routine biopsy and could be streamlined for a central laboratory with results obtainable in 2 days. Furthermore, this procedure works successfully on formalin-fixed paraffin-embedded sections, thus allowing retrospective studies of biopsies or autopsy materials.

Cancer biology for individualized therapy: correlation of growth fraction index in native-state histoculture with tumor grade and stage

There is a need for individualization of all aspects of cancer therapy. Because of significant heterogeneity within a tumor class, there is a need to develop an in vitro test to accurately gauge tumor aggressiveness. Such a measurement would greatly aid treatment decision making. Current methodologies such as flow cytometry, which lacks unambiguous interpretation of cell-proliferative data, and determination of the thymidine-labeling index, which measures nucleotide uptake in a nonphysiological state, have not reproducibly attained this goal. We have developed an in vitro native-state three-dimensional gel-supported histoculture system that allows the growth of all human solid tumor types for relatively long time periods. The native-state system was used to identify the percent of cells capable of incorporating [3H]thymidine over a 4-day period, which we term the growth fraction index (GFI). We have compared the ability of cancer tissue to proliferate in native-state culture to the stage and histological grade of four major types of human carcinomas: breast, ovarian, colon, and lung. Eighty percent of tumor explants could be evaluated, even when sent from across the country. We have determined that the GFI correlates with tumor stage and grade for breast and ovarian carcinoma. In colon carcinoma, there is a trend toward higher GFIs in tumors of more advanced stage and grade. In non-small cell lung carcinomas, GFI, stage, and grade do not correlate. These results suggest the applicability of gel-supported three-dimensional native-state histoculture for prognostic purposes in patients with breast and ovarian cancers and demonstrate the clinical relevance of the native-state histoculture system.