ANTICANCER RESEARCH 26: 1765-1772 (2006)

Ex Vivo Assessment of -induced Apoptosis and Associated Molecular Changes in Patient Tumor Samples

FARZANEH PIRNIA1, STEFFEN FRESE2, BEAT GLOOR3, MICHEL A. HOTZ4, ALEXANDER LUETHI1, MATHIAS GUGGER5, DANIEL C. BETTICHER1 and MARKUS M. BORNER1

1Institute of Medical Oncology, 2Clinic for Thoracic , 3Department of Visceral and Transplantation Surgery and 4Nose and Throat Surgery, Institute of Pathology University of Bern, Inselspital, Bern, Switzerland

Abstract. Background: There are inherent conceptual problems It is notoriously difficult to study the molecular effects of in investigating the pharmacodynamics of drugs . anticancer drug treatment in vivo. To date, the most common One of the few possible approaches is serial biopsies in patients. approach is to perform serial biopsies in tumor patients However, this type of research is severely limited by undergoing treatment. Such studies are associated with methodological and ethical constraints. Materials and Methods: discomfort, costs and significant morbidity, even mortality. A modified 3-dimensional tissue culture technique was used to Thus, patients are not motivated to participate and Ethics culture human tumor samples, which had been collected during Committees are very reluctant to support this kind of routine cancer operations. Twenty tumor samples of patients with research. An approach to circumvent these problems is ex vivo non-small cell lung cancer (NSCLC) were cultured ex vivo for tissue culture. Tumor tissue is obtained from patients during 120 h and treated with mitomycin C, taxotere and cisplatin. The routine operation and is then processed . Hoffman et cytotoxic activity of the anticancer agents was quantified by al. pioneered this approach by establishing a protocol for ex assessing the metabolic activity of treated tumor cultures and vivo cultivation of 3-dimensional tissue samples (1-3). Tumors various assays of apoptosis and gene expression were performed. are processed into small cubes, which can be grown on a Results: The proliferative activity of the tissue was maintained in collagen gel support. Three-dimensional culture methods have culture as assessed by Ki-67 staining. Mitomycin C, cisplatin and been compared with 2-dimensional culture methods, which taxotere reduced the metabolic activity of the tumor tissue still are the most frequently used culture method in cancer cultures by 51%, 29% and 20%, respectively, at 120 h. The research. Two-dimensional techniques have the decrease in metabolic activity corresponded to the induction of advantage of simple handling and high reproducibility. apoptosis as demonstrated by the typical morphological changes, However, 2-dimensional and 3-dimensional culture techniques such as chromatin condensation and nuclear fragmentation. In have shown critical biological differences. The most obvious addition, activated caspase-3 could be verified in apoptotic cells are differences in the expression of genes, which are involved by immunohistochemistry. To verify functional aspects of in attachment, adhesion and the organization of the apoptosis, the induction of chemotherapy-induced cell death was 3-dimensional tissue architecture. Even more importantly for inhibited with the caspase inhibitor z-VAD.fmk. RNA was cancer research, the spatial organization of cells has a clear extracted from the tissue cultures after 120 h of ex vivo drug impact on their sensitivity to anticancer drugs and other treatment and was of sufficient quality to allow quantitative PCR. therapeutic manipulations (4-7). Thus, 3-dimensional tissue Conclusion: The 3-dimensional ex vivo culture technique is a cultures seem to reflect more appropriately the in vivo patient useful method to assess the molecular effects of pharmacological situation than the commonly used 2-dimensional monolayer interventions in human cancer samples in vitro. This culture cell cultures. technique could become an important tool for drug development Practicability is an important requirement before a new and for the prediction of in vivo drug efficacy. culture technique can be widely used in cancer research. A critical issue in this context is how to assess cytotoxicity as a major end-point for therapeutic drug action in culture. To date, viability in 3-dimensional tissue cultures has mainly been Correspondence to: Markus M. Borner, MD, Institute of Medical measured with the 3-(6)-2, 5-diphenyl tetrazolium bromide Oncology, Inselspital, 3010 Bern, Switzerland. Tel: +41 (31) 632 (MTT) dye assay or with the (8) thymidine incorporation 84 42, Fax: +41 (31) 382 41 19, e-mail: [email protected] assay. Both assays are very labor intensive and an additional Key Words: Apoptosis, ex vivo, non-small cell lung cancer, important disadvantage of the (8) thymidine incorporation quantitative PCR. assay is the requirement for radioactivity. Thus, the possibility

0250-7005/2006 $2.00+.40 1765 ANTICANCER RESEARCH 26: 1765-1772 (2006) of using Alamar Blueì to assess cytotoxic drug action in assay measurement and the tissue was fixed in 4% paraformaldehyde. 3-dimensional tissue cultures was explored. Alamar Blueì is The metabolic activity of the tissue was measured with Alamar Blueì a non-fluorescent compound, which yields a fluorescent in a Spectramax Gemini Fluorometer (Molecular Devices, Sunnyvale, CA, USA) at 560 nm excitation wavelength and 590 nm emission product after metabolic reduction by living cells. In addition to wavelength. Less than 5% reduction of metabolic activity was avoiding high costs and the use of radioactivity, this method considered as non-responsive. allows cytotoxicity to be measured without processing the tissue cultures, since the dye can be directly added to the Hoechst staining. Two-Ìm paraffin sections were placed on SuperFrost culture medium. Alamar Blueì is not cytotoxic and tissue Plus slides (Roth, Karlsruhe, Germany). The slides were dewaxed and cultures can be further incubated in its presence (8-11). stained with Hoechst 33342, as previously described (15). Continued growth leads to a reduced environment (pink Immunohistochemistry of activated caspase-3. The paraffin sections color), while inhibition of growth maintains an oxidized were dewaxed and dehydrated. For antigen unmasking, the tissue environment (blue color; see Figure 1B). sections were boiled in 10 mM sodium citrate buffer (pH 6.0) for The molecular program of apoptosis has been identified as 1 min at full power followed by 9 min at medium power in a a major determinant of chemotherapy responsiveness in microwave oven and were then cooled and washed with PBS. After cancer. In addition to a great diversity of individual drug- blocking endogenous peroxides with 1% H2O2 for 10 min and 5% target interactions, all anticancer agents achieve their horse serum for 1 h, the slides were either incubated with activated therapeutic goal of inducing cell death by exhibiting the caspase-3 antibody (Cell Signaling Technology, MA, USA) (1:50 dilution) or Ki-67 polyclonal antibody (Santa Cruz, CA, USA) for stereotypical morphological features of programmed cell 1 h. For detection of the immunosignal, the ABC Reagent death (apoptosis) (12-14). Different aspects of apoptosis (Vectastain ABC kit, Vector Laboratories, Inc., Burlingame, CA, induction by chemotherapy drugs were examined in USA) was used. The sections were counterstained with hematoxylin. 3-dimensional ex vivo cultures of non-small cell lung cancer The cells were examined on a Leica DMRB microscope using a 40x (NSCLC) patient samples. The goal of this work was to magnification lens. establish this culture technique for use in cancer research. Caspase inhibitor. The caspase inhibitory peptide benzyloxycarbonyl- Materials and Methods Val-Ala-Asp- fluoromethyl ketone (z-VAD.fmk) was obtained from Bachem (Bubendorf, Switzerland). The caspase inhibitors were used The histoculture drug response assay. Tumor tissue from 20 patients at a concentration of 100 ÌM 2 h prior to chemotherapy exposure. (6 female, 14 male) with NSCLC was collected upon resection of the primary tumor. All patients gave their written consent to be included RNA extraction and reverse transcription. Total RNA was extracted in this study. Gelfoam (Pharmacia & Upjohn, Kalamazoo, MI, USA) and cDNA synthesis was performed as previously described (16). pieces of 1 cm2 were placed into each well of 12-well plates. The gelfoam was allowed to soak for 2 h with culture medium (RPMI Real-time quantitative RT-PCR. Real-time PCR was performed as 1640 medium) containing 10% (v/v) heat-inactivated fetal bovine previously described (16). Briefly, real-time PCR was used for serum and 100 units/ml antibiotic antimycotic solution from Sigma, quantitation of p73 N-terminal variants, while 7S RNA was used for Switzerland. NSCLC tissue samples were taken from the operating the internal standard. Each PCR was carried out in a total volume of room immediately after resection and placed in transport culture 25 Ìl containing cDNA reverse-transcribed from 25 ng and 0.375 ng medium. Areas containing tumor were identified by routine total RNA for p73 and 7S, respectively. Dual-labelled (FAM/TAMRA) histological examinations. The specimens were dissected into 2 to gene-specific probes and TaqMan Universal PCR Master Mix 3 mm cubes. About 40-45 Ìg of tumor tissue was placed on top of the (Applied BioSystems, Rotkreuz, Switzerland) were used for the PCR. gelfoam in the 12-well plates, which were incubated at 37ÆC with 5% ¢N p73 was amplified using 5'GGAGATGGGA AAAGCG AAAAT-3' as the forward primer, 5'CTCTCCCGCTC GG-TCCAC- CO2 overnight. After 24 h, the chemotherapeutic drugs were added directly to the 3' as the reverse primer (both 300 nM) and a ¢N p73 probe respective well and the tissue cultures were incubated for the 5'CAAACGG-CCCGCATGTTCCC-3' (150 nM). For TA p73, the indicated time period (Figure 1A). Taxotere (1 Ìg/ml) was purchased primers were forward 5'-GC-ACCACGTTTGAGCACCTC-3', from Rhöne-Poulenc Rorer (Thalwil, Switzerland), cisplatin reverse 5'-TAATGAGGTGGTGGGCGGA-3' (both 300 nM), and (10 Ìg/ml) and mitomycin C (MMC) (5 Ìg/ml) were purchased from the probe was 5'-TTCGACCTTCCCCAGTCA AGCCG-3' (150 nM). Bristol-Myers Squibb (Baar, Switzerland). The drugs were used at For 7S RNA, the primers and probe were: forward 5'-ACCACCA clinically relevant concentrations. For each treatment condition, 10% GGTTGCCTAAGGA-3’ reverse 5'CACGGGAG TTTTGACCTG (v/v) Alamar Blueì (Serotech, Raleigh, NC, USA) was added to CT-3', probe 5'TGAACCGG CCCAGGT-CGGAAAC-3' (300 nM measure the metabolic activity of the tissue (Figure 1B). Alamar each). All measurements were performed twice and the arithmetic Blueì is reduced by cellular mitochondrial enzymes. The product has mean was used for further calculations. a lower peak absorbance value and a substantially enhanced fluorescence, which can be measured spectroscopically in cell medium Results without further processing. Living metabolically active cells lead to a reduced environment (pink color), while inactive or dead cells maintain an oxidized environment (blue color). Following 3 to 4 h of Ex vivo tissue cultures. All 20 NSCLC tumor tissue samples Alamar Blueì incubation, the medium was collected for metabolic could be cultivated without super-infection or other culture

1766 Pirnia et al: Ex Vivo Assessment of Chemotherapy-induced Apoptosis in Tumor Samples

Figure 2. Cytotoxic effect of various anticancer drugs on 3-dimensional tissue cultures from NSCLC tissue patient samples. (A) Continuous Figure 1. The histoculture drug response assay in NSCLC tissue. (A) treatment with 5 Ìg/ml MMC, 10 Ìg/ml cisplatin and 0.1 Ìg/ml taxotere Human lung cancer tissue was treated with MMC (5 Ìg/ml) in the absence for the indicated duration. (B) Similar experiments as above in the and presence of 100 ÌM zVAD.fmk for 120 h. (B) Same as A in the absence and presence of 100 ÌM zVAD.fmk for 5 days. Data are the mean presence of Alamar Blueì. The samples were run as duplicates. of patients in each treatment group.

problems. A typical example of a culture plate is shown in death had to be verified. Apoptosis is defined by the Figure 1A. The tumor samples were incubated with various occurrence of a stereotypical phenotype, including cell anticancer agents for 120 h. The number of drugs tested was shrinkage, chromatin condensation and nuclear fragmentation. dependent on the amount of available tumor tissue. Untreated To demonstrate the induction of apoptotic cell death, controls were run in parallel on the same 12-well plates. The chemotherapy-treated paraffin sections of tumor tissue after viability of the tissue cultures was assessed by the Alamar various treatment conditions were stained with the DNA- Blueì fluorometric dye assay (Figure 1B). The viability of the intercalating Hoechst 33342 dye (15). The untreated controls tumor tissues was decreased by 51% (median value) in the showed an intact nuclear structure with this staining (Figure samples from 16 different tumors (100%), which were treated 3A). The treated samples underwent nuclear condensation with 5 Ìg/ml MMC (Figure 2A). In comparison, only 56% typical of apoptosis. As a representative example, an MMC- (five out of nine) of the NSCLC samples responded to treated tumor sample is depicted in Figure 3B. The cisplatin (10 Ìg/ml). The median reduction of viability by morphological apoptotic changes were similar to those cisplatin was 29% after 120 h. Four out of eight tumor obtained with the other drugs and the extent of apoptotic samples (50%) responded to taxotere (10 Ìg/ml) with a changes corresponded to the degree of metabolic inhibition. median reduction of metabolic activity by 20%. However, an exact quantification of apoptotic nuclei was not possible due to fragmentation of nuclei, sampling error and Ex vivo induction of apoptosis with anticancer agents in human heterogeneous distribution of the morphological changes lung tumor tissue. As shown in Figure 2A, incubation with throughout the tissue cube depending on the drug penetration. various anticancer agents led to a decrease in metabolic activity Apoptosis is an active process, executed by caspases. The in ex vivo tumor tissue cultures. Whether the decrease in broad-spectrum, cell membrane-permeable caspase inhibitor metabolic activity was due to the induction of apoptotic cell z-VAD.fmk was used to block the occurrence of apoptosis in

1767 ANTICANCER RESEARCH 26: 1765-1772 (2006) treated tumor tissues. As shown in Figure 3C, apoptotic cell death could be largely prevented by co-administering z-VAD.fmk with MMC. As shown in Figure 2B, metabolic activity could also be restored by co-administering the apoptosis inhibitor z-VAD.fmk with the respective anticancer agent. These findings suggest that the decrease of metabolic activity in this assay was an expression of the induction of apoptosis.

Activation of caspase-3 with chemotherapy. The characteristic morphological features of apoptosis are executed by a family of aspartate-specific cysteine proteases (caspases), which can be activated by various apoptogenic signals (17, 18). In the classic model, caspases are divided into initiator caspases (such as caspases-8, -9) and executioner caspases (such as caspases-3, -6, -7) according to their function and their sequence of activation (17, 18). Caspase-3 seems to play a central role in chemotherapy-induced apoptosis. In the process of activation, procaspase-3 is cleaved to caspase-3 fragment, which can be stained with a specific antibody. The negative staining for untreated tissue is represented in Figure 4A, while MMC-treated cells exhibited strong perinuclear staining for caspase-3 fragment as shown in Figure 4B. This figure clearly shows that in ex vivo-treated tissue cultures, caspase-3 is involved in chemotherapy-induced apoptosis. As a proof of principle, the prevention of caspase-3 activation could also be demonstrated by co-administering the caspase inhibitor z-VAD.fmk with MMC (Figure 4C).

Quantitative analysis of TA p73, ¢N p73. p73 has significant homology to p53. The TA p73 gene encodes two different proteins, full-length TA p73 and the truncated protein ¢N p73. TA p73 acts as a tumor suppressor and ¢N p73 seems to behave as an oncogene (19). It has been described by Lunghi et al. that chemotherapy promotes the accumulation of endogenous TA p73 and the reduction of ¢N p73 (20). Therefore, the expression levels of TA p73 and ¢N p73 were quantified in 3-dimensional tissue cultures from NSCLC patient samples before and after treatment with MMC by quantitative PCR. A 3-fold increase in TA p73 and a 10-fold decrease in ¢N p73 expressions were found after MMC treatment compared to untreated tumor tissue in a patient Figure 3. Drug-induced apoptosis in NSCLC tissue tissue cells. NSCLC (Figure 5). This result suggested that anticancer therapy can tissue was treated with 5 Ìg/ml MMC for 120 h and stained with Hoechst induce similar changes in the expression of ¢N p73 and 33342. (A) Untreated NSCLC tissue, (B) MMC-treated NSCLC tissue TA p73 levels in ex vivo tissue cultures as observed in other without, or with100 ÌM z-VAD-fmk (C), magnification, x 40. experimental systems. Our results also clearly showed that it is possible to isolate good quality RNA from ex vivo tissue cultures, allowing for further processing for molecular biology is a nuclear protein that is expressed in proliferating cells techniques, such as Northern blots or quantitative PCR. and has been used as a marker for cell proliferation in solid tumors, as well as in some hematological malignancies (21). Demonstration of proliferative activity by Ki-67 staining. To As shown in Figure 6, Ki-67 nuclear staining was observed demonstrate that tumor tissue is still capable of proliferation in human lung cancer tissue after 5 days in culture, in ex vivo tissue cultures, the proliferative activity was suggesting that even after this period of culture, the tumor assessed by immunohistochemical staining for Ki-67. Ki-67 tissue is still able to proliferate ex vivo.

1768 Pirnia et al: Ex Vivo Assessment of Chemotherapy-induced Apoptosis in Tumor Samples

Figure 4. Immunohistochemistry of activated caspase-3. (A) Untreated human lung cancer tissue, treated with MMC (5 Ìg/ml) in the absence (B) or presence of 100 ÌM zVAD.fmk for 120 h (C), magnification, x 40.

1769 ANTICANCER RESEARCH 26: 1765-1772 (2006)

human body, tissue has a 3-dimensional architecture and the biological significance of stromal cells for the function of normal organs and tumors has only recently begun to be appreciated. The 3-dimensional architecture comes with interactional processes, such as attachment, adhesion, spatial orientation and the impact of gravity, which all influence the biology of the involved cells (4-7). This context is specially relevant for studies on therapeutic drug action in oncology, since adhesion and attachment have been shown to directly influence the apoptosis threshold of involved cells (4, 32-39). Most anticancer drugs achieve their antitumor effect by inducing apoptosis (12-14), while factors, which set the apoptosis threshold higher, can contribute to multi-drug resistance (36, 38-40). There are studies supporting the notion that higher drug levels are needed to achieve cytotoxic action in 3-dimensional tissue cultures as compared to the more conventional cell cultures or cell dispersions (2, 5, 6, 31). Thus, 3-dimensional cultures could be the more clinically relevant system to investigate drug effects on tumor tissue. It was shown here that the 3-dimensional culture technique allows for treatment of human tissue ex vivo and for the investigation of drug-related molecular effects over time. The typical apoptosis hallmarks could be induced by cytotoxic Figure 5. Real-time RT-PCR experiments for TA p73 and ¢N p73 expressions. Human lung cancer tissue was treated without or with MMC drugs and, as a proof of concept, apoptosis induction and (5 Ìg/ml) for 120 h. The transcript numbers were normalized using 7S measures of cytotoxicity could be inhibited by co- ribosomal RNA as a house-keeping gene. X-fold induction was calculated administering the caspase inhibitory peptide z-VAD.fmk. using the normalized transcript number. These results add further evidence that anticancer drugs also induce apoptosis in vivo and that the cancer cell has to participate actively in its own demise by providing functional Discussion apoptosis machinery. In addition to morphological studies, DNA and RNA can be extracted from the cultured tissue Several lines of reasoning support the idea that 3-dimensional even after 5 days of incubation and the biological molecules ex vivo tissue culture is a very useful tool for drug testing. are of sufficient quality to perform Northern blots (data not There is the practicability and patient convenience issue, since shown) or quantitative PCR analyses. We have recently used ex vivo treatment of patient tissue, which has been sampled this method successfully for reverse phase proteomics during routine surgery, allows for an assessment of the microarrays (manuscript in preparation). The simplicity of the pharmacodynamics of the drug over several days. In our cytotoxicity end-point assessment with Alamar Blueì makes experience with NSCLC tissue, there were no super-infections this method attractive for clinical studies. This dye can be if cultures were maintained for 5 days, as corroborated by added without further manipulation to the culture medium, other investigations (1, 2, 22-25) and it seems that this culture which can be sampled at various time-points for cytotoxicity period can even be extended (23, 26). However, it is possible measurements by fluorometry. This is a simplification of the that tissue from sources such as colon cancer or head and method as originally described by Hoffman et al. (1, 2) without neck tumors is more difficult to culture since these sites are sacrificing accuracy (8-11). typically colonized by bacteria and fungi. Other authors have The practical applications of 3-dimensional culture include reported culture problems in about 10-30% of such samples the development of a drug-response assay that correlates with (23, 27-30). Our experiments demonstrated that untreated drug sensitivity and the survival of cancer patients. Some human NSCLC tissue was still viable after 5 days of culture interesting results have already been published using the and was able to proliferate, as described for other tumors (1). original method by Hoffman et al. (2, 25, 30, 40, 41), but the Another crucial point is the biological relevance of this data are preliminary and the method has proven too time- system (31). Many important findings in cancer research are consuming for broad application. Many different drug- still of questionable clinical relevance, since they have only response assays have been explored, but have fallen out of been demonstrated in highly artificial systems, such as animal favor because of the lack of clear results and well designed models or in 2-dimensional cell cultures. However, in the clinical studies (42). Different methodological and ethical

1770 Pirnia et al: Ex Vivo Assessment of Chemotherapy-induced Apoptosis in Tumor Samples

Figure 6. Demonstration of proliferative activity of NSCLC tissue by Ki-67 staining. The tissue was cultured for 5 days and stained with Ki-67, magnification, x 40. constraints limit the design of such studies, despite their 2 Vescio RA, Redfern CH, Nelson TJ, Ugoretz S, Stern PH and clinical relevance. Drug sensitivity assays may become more Hoffman RM: In vivo-like drug responses of human tumors important in ensuing years as an increasing number of new growing in three-dimensional gel-supported primary culture. Proc Natl Acad Sci USA 84(14): 5029-5033, 1987. active drugs, enter the therapeutic armamentarium. It 3 Hoffman RM: The three-dimensional question: can clinically remains to be shown that the 3-dimensional culture technique relevant tumor drug resistance be measured in vitro? Cancer can make a difference in comparison to the thus far Metastasis Rev 13(2): 169-173, 1994. unsuccessfully explored in vitro chemosensitivity assays. 4 Nakamura T, Kato Y, Fuji H, Horiuchi T, Chiba Y and Tanaka Having a system on hand, which mimics the treatment of real K: E-cadherin-dependent intercellular adhesion enhances patients more closely than other methods, will hopefully help chemoresistance. Int J Mol Med 12(5): 693-700, 2003. the development of novel therapeutic strategies and aid the 5 Green SK, Frankel A and Kerbel RS: Adhesion-dependent multicellular drug resistance. Anticancer Drug Des 14(2): 153- prediction of drug sensitivity. 168, 1999. 6 Furukawa T, Kubota T, Watanabe M, Kase S, Saikawa Y, Acknowledgements Nishibori H et al: Increased drug resistance of cultured human cancer cell lines in three-dimensional cellular growth assay Oncoswiss, the Bern Cancer League and the Annemarie Waefler using collagen gel matrix. J Surg Oncol 49(2): 86-92, 1992. Foundation are thanked for their support. 7 Kerbel RS, St. Croix B, Florenes VA and Rak J: Induction and reversal of cell adhesion-dependent multicellular drug References resistance in solid breast tumors. Hum Cell 9(4): 257-264, 1996. 1 Hoffman RM, Connors KM, Meerson-Monosov AZ, Herrera H 8 Ahmed SA, Gogal RM Jr and Walsh JE: A new rapid and and Price JH: A general native-state method for determination of simple non-radioactive assay to monitor and determine the proliferation capacity of human normal and tumor tissues in vitro. proliferation of lymphocytes: an alternative to [3H]thymidine Proc Natl Acad Sci USA 86(6): 2013-2017, 1989. incorporation assay. J Immunol Methods 170(2): 211-224, 1994.

1771 ANTICANCER RESEARCH 26: 1765-1772 (2006)

9 Hamid R, Rotshteyn Y, Rabadi L, Parikh R and Bullock P: correlation of growth fraction index in native-state histoculture Comparison of Alamar Blue and MTT assays for high through- with tumor grade and stage. Proc Natl Acad Sci USA 87(2): 691- put screening. Toxicol In Vitro 18(5): 703-710, 2004. 695, 1990. 10 Gloeckner H, Jonuleit T and Lemke HD: Monitoring of cell 28 Furukawa T, Kubota T, Watanabe M, Kase S, Takahara T, viability and cell growth in a hollow-fiber bioreactor by use of the Yamaguchi H et al: Chemosensitivity testing of clinical dye Alamar Blue. J Immunol Meth 252(1-2): 131-138, 2001. gastrointestinal using histoculture and the MTT end- 11 Byth HA, McHunu BI, Dubery IA and Bornman L: Assessment point. Anticancer Res 12(5): 1377-1382, 1992. of a simple, non-toxic Alamar Blue cell survival assay to monitor 29 Robbins KT, Varki NM, Storniolo AM, Hoffman H and Hoffman tomato cell viability. Phytochem Anal 12(5): 340-346, 2001. RM: Drug response of head and neck tumors in native-state 12 Johnstone RW, Ruefli AA and Lowe SW: Apoptosis: a link histoculture. Arch Otolaryngol Head Neck Surg 117(1): 83-86, between cancer genetics and chemotherapy. Cell 108(2): 153-164, 1991. 2002. 30 Singh B, Li R, Xu L, Poluri A, Patel S, Shaha AR et al: Prediction 13 Dixon SC, Soriano BJ, Lush RM, Borner MM and Figg WD: of survival in patients with head and neck cancer using the Apoptosis: its role in the development of malignancies and its histoculture drug response assay. Head Neck 24(5): 437-442, 2002. potential as a novel therapeutic target. Ann Pharmacother 31(1): 31 Hoffman RM: To do tissue culture in two or three dimensions? 76-82, 1997. That is the question. Stem Cells 11(2): 105-111, 1993. 14 Borner MM, Myers CE, Sartor O, Sei Y, Toko T, Trepel JB et al: 32 Barzanti F, Zoli W, Susino MD, Ricotti L, Tesei A, Papa S et al: Drug-induced apoptosis is not necessarily dependent on Simultaneous determination of apoptosis and surface antigen macromolecular synthesis or proliferation in the p53-negative expression in tumor adherent cells. J Biol Regul Homeost Agents human prostate cancer cell line PC-3. Cancer Res 55(10): 2122- 15(4): 359-365, 2001. 2128, 1995. 33 Brouwer RE, Hoefnagel J, Borger van Der Burg B, Jedema I, 15 Pirnia F, Schneider E, Betticher DC and Borner MM: Mitomycin Zwinderman KH, Starrenburg IC et al: Expression of co- C induces apoptosis and caspase-8 and -9 processing through a stimulatory and adhesion molecules and chemokine or apoptosis caspase-3 and Fas-independent pathway. Cell Death Differ 9(9): receptors on acute myeloid leukaemia: high CD40 and CD11a 905-914, 2002. expression correlates with poor prognosis. Br J Haematol 115(2): 16 Grob TJ, Novak U, Maisse C, Barcaroli D, Luthi AU, Pirnia F et 298-308, 2001. al: Human delta Np73 regulates a dominant negative feedback 34 Damiano JS, Hazlehurst LA and Dalton WS: Cell adhesion- loop for TAp73 and p53. Cell Death Differ 8(12): 1213-1223, 2001. mediated drug resistance (CAM-DR) protects the K562 chronic 17 Strasser A, Harris AW, Huang DC, Krammer PH and Cory S: myelogenous leukemia cell line from apoptosis induced by Bcl-2 and Fas/APO-1 regulate distinct pathways to lymphocyte BCR/ABL inhibition, cytotoxic drugs, and gamma-irradiation. apoptosis. EMBO J 14(24): 6136-6147, 1995. Leukemia 15(8): 1232-1239, 2001. 18 Thornberry NA and Lazebnik Y: Caspases: enemies within. 35 Kim JB, Stein R and O'Hare MJ: Three-dimensional in vitro Science 281(5381): 1312-1316, 1998. tissue culture models of breast cancer – a review. Breast Cancer 19 Melino G, De Laurenzi V and Vousden KH: p73: friend or foe in Res Treat 85(3): 281-291, 2004. tumorigenesis. Nat Rev Cancer 2(8): 605-615, 2002. 36 Morin PJ: Drug resistance and the microenvironment: nature and 20 Lunghi P, Costanzo A, Levrero M and Bonati A: Treatment with nurture. Drug Resist Updat 6(4): 169-172, 2003. arsenic trioxide (ATO) and MEK1 inhibitor activates the p73- 37 Truong T, Sun G, Doorly M, Wang JY and Schwartz MA: p53AIP1 apoptotic pathway in leukemia cells. Blood 104(2): 519- Modulation of DNA damage-induced apoptosis by cell adhesion 525, 2004. is independently mediated by p53 and c-Abl. Proc Natl Acad Sci 21 Gore SD, Weng LJ and Burke PJ: Validation of flow-cytometric USA 100(18): 10281-10286, 2003. determination of Ki67 expression as a measure of growth factor 38 Kraus AC, Ferber I, Bachmann SO, Specht H, Wimmel A, Gross response in acute myelogenous leukemia. Exp Hematol 21(13): MW et al: In vitro chemo- and radio-resistance in small cell lung 1702-1708, 1993. cancer correlates with cell adhesion and constitutive activation of 22 Vescio RA, Connors KM, Bordin GM, Robb JA, Youngkin T, AKT and MAP kinase pathways. Oncogene 21(57): 8683-8695, 2002. Umbreit JN et al: The distinction of small cell and non-small cell 39 Rintoul RC and Sethi T: Extracellular matrix regulation of drug lung cancer by growth in native-state histoculture. Cancer Res resistance in small-cell lung cancer. Clin Sci (Lond) 102(4): 417- 50(18): 6095-6099, 1990. 424, 2002. 23 Griffon-Etienne G, Merlin JL and Marchal C: Evaluation of 40 Tanino H, Oura S, Hoffman RM, Kubota T, Furukawa T, Taxol in head and neck squamous carcinoma multicellular tumor Arimoto J et al: Acquisition of multidrug resistance in recurrent spheroids. Anticancer Drugs 8(1): 48-55, 1997. breast cancer demonstrated by the histoculture drug response 24 Freeman AE and Hoffman RM: In vivo-like growth of human assay. Anticancer Res 21(6A): 4083-4086, 2001. tumors in vitro. Proc Natl Acad Sci USA 83(8): 2694-2698, 1986. 41 Hoffman RM: In vitro sensitivity assays in cancer: a review, 25 Kubota T, Sasano N, Abe O, Nakao I, Kawamura E, Saito T et al: analysis, and prognosis. J Clin Lab Anal 5(2): 133-143, 1991. Potential of the histoculture drug-response assay to contribute to 42 Schrag D, Garewal HS, Burstein HJ, Samson DJ, Von Hoff DD cancer patient survival. Clin Cancer Res 1(12): 1537-1543, 1995. and Somerfield MR: American Society of Clinical Oncology 26 Li LN, Margolis LB and Hoffman RM: Skin toxicity determined Technology Assessment: chemotherapy sensitivity and resistance in vitro by three-dimensional, native-state histoculture. Proc Natl assays. J Clin Oncol 22(17): 3631-3638, 2004. Acad Sci USA 88(5): 1908-1912, 1991. 27 Vescio RA, Connors KM, Youngkin T, Bordin GM, Robb JA, Received January 23, 2006 Umbreit JN et al: Cancer biology for individualized therapy: Accepted March 1, 2006

1772