Metastatic Soft Tissue Sarcoma Chemotherapy: An Opportunity for Personalized Medicine

Transcription

1 Clinically validated predictive tests for systemic therapy response in soft tissue sarcoma are still being sought. Jeffrey Hessing. Maison Ancienne. Oil on canvas, 31¾ 39. Metastatic Soft Tissue Sarcoma Chemotherapy: An Opportunity for Personalized Medicine Damon Reed, MD, and Soner Altiok, MD, PhD Background: Soft tissue sarcoma (STS) includes biologically and histologically diverse mesenchymal tumors that are relatively chemotherapy-resistant compared with other sarcoma subtypes. Methods: The authors discuss the clinical challenges frequently encountered by medical oncologists and review the literature for predictive strategies to systematically approach chemotherapy decision making. Results: There are no clinically validated predictive tests for chemotherapeutic response or resistance in STS. Clinical features including histology, stage, and patient age are currently used to guide therapy decisions in STS. Conclusions: A method to predict response or resistance to chemotherapy, utilizing both targeted and conventional agents, would be beneficial in reducing toxicity and improving response rates for patients with STS and also in designing clinical trials for this disease. Introduction Soft tissue sarcoma (STS) encompasses a range of over 40 histologic diagnoses with varying biology, from translocation-defined entities to diseases with complex karyotypes. 1-4 Roughly 10,000 new cases are diagnosed in the United States per year across all ages, comprising 1% of adult cancers and 7% of childhood cancers. While localized resected disease can often be cured, controversy exists regarding optimal management of tumors in younger patients with high-grade disease or larger From the Departments of Sarcoma (DR) and Pathology (SA) at the H. Lee Moffitt Cancer Center & Research Institute, Tampa, Florida. Submitted June 17, 2010; accepted September 23, Address correspondence to Damon Reed, MD, Moffitt Cancer Center, Sarcoma Department, Magnolia Drive, FOB 1, Tampa, FL Damon.Reed@moffitt.org The authors receive support from the Sarcoma Foundation of America, V Foundation, and the Moffitt Cancer Center Sarcoma Foundation. No significant relationship exists between the authors and the companies/organizations whose products or services may be referenced in this article. tumors. Prognosis is significantly worse when tumors cannot be controlled by surgery and radiation therapy or in the setting of metastatic disease, and chemotherapy is the primary modality of treatment. Because of the poor overall survival of 20% to 25% at 2 years, as well as the short median survival of 12 months and the rarity of these tumors, there is a need for novel approaches for therapy. 5 Due to the biologic diversity, it is unlikely that a single agent or a combination of agents would be successful across the spectrum of diagnoses. Chemotherapy in STS Historically, doxorubicin-based therapy has been the standard first-line agent in the treatment of STS when chemotherapy is indicated, typically in metastatic or unresectable disease. While combination therapy, mainly with ifosfamide, has improved response rates, toxicities have increased with no clear corresponding benefit in overall survival Response rates to front-line chemotherapy vary, ranging from 10% to 46% with single agents and with combinations of chemotherapy. 5,9-19 Confounding this even further is the spectrum of response rates 188 Cancer Control July 2011, Vol. 18, No. 3

2 between histologies. 4,16 In addition, patient age affects analyses; improved responses in younger patients have been reported Finally, specific subtypes may benefit from agents outside the typical approach of anthracyclines and alkylators (eg, taxanes for angiosarcoma and trabectedin for myxoid/round cell liposarcomas). 25,26 Tyrosine kinase inhibitors have revolutionized the management of gastrointestinal stromal tumors (GISTs), a subtype of STS, showing that significant clinical benefit is possible among the diagnoses currently categorized as STS if the correct therapy can be matched to the tumor biology. 27 Interpretation between trials is difficult in part due to the heterogeneity of diagnoses and ages among cohorts and has contributed to conflicting results between trials. Neither a systematic method to guide chemotherapy for an individual patient nor a clear national standard of care has yet been developed, thus necessarily leading to a multifactorial, patient-specific analysis to determine whether or not to offer therapy. This is particularly complicated in second-line therapies, where many agents are available with similar low response rates, and the short median survival of these patients does not always allow for multiple treatment attempts. Response rates are so low in the second-line setting that progression-free survival (PFS) has become the standard endpoint in clinical trials, with PFS at 3 months above 40% demonstrating activity and below 20% demonstrating inactivity. 28 Ifosfamide is a standard second-line agent, with dosedependent activity ranging from 10% to 40% with particular activity in synovial cell sarcoma. 12,18,19 Gemcitabine and docetaxel combination chemotherapy has demonstrated particularly good response rates in leiomyosarcoma, especially uterine leiomyosarcoma, and has become front-line therapy in our institution for this malignancy. Response rates of 16% to 43% have been reported in STS series. 13,15-17 Vinorelbine has particular activity in rhabdomyosarcoma, having been studied particularly in the pediatric population; it also has a 6% response rate in heavily pretreated STS. 11,14,29 Many agents beyond the scope of this article are being explored in STS. These include agents with a wide range of activity, such as cytotoxic and targeted agents. Trabectedin and ifosfamide derivatives are currently being studied in phase III studies, and mtor inhibitors have been studied in maintenance, with approval pending. Insulin-like growth factor (IGF) pathway inhibitors, proteasome inhibitors, cell cycle and growth modulators, kinase inhibitors and others are being explored in earlier phase trials, at times with corresponding biomarker development. The vast numbers of available agents and numerous histologic diagnoses complicate study design, which can lead to overall negative conclusions and also overlook clinical benefit in a particular subtype of STS. A means to enrich responders and exclude nonresponders in early-phase trials for STS is needed. Assay-Guided Therapy Prospective determination of antibiotic sensitivity and resistance has been the standard of care in infectious diseases for many years. In contrast, due to the lack of reproducible and predictive assays, treatment protocols of cancer patients have been designed according to tumor histology rather than to the tumor s sensitivity to a given chemotherapeutic agent. Therefore, development of novel drug sensitivity studies to predict patient response before initiation of therapy would help to design the most efficient therapeutic regimen for individual patients while sparing others who are not likely to benefit from the given therapy. A formal analysis conducted by both the American Society of Clinical Oncology and the Blue Cross and Blue Shield Association s Technology Evaluation Center concluded that predictive assays were not ready for routine clinical use. 30,31 However, both studies noted that versions of these assays that meet specific parameters might improve prediction of patient response to therapy. Reasons for specific problems with previous assays included the low percentage of successfully completed assays that yielded clinically useful results, rare interpretations that differed from standard of care, trial design that did not include choices among several treatment options, technical complexity that made application beyond a single laboratory/institution unlikely, and the long time needed to study completion, thus delaying therapy. 30,31 Minimum criteria that have been proposed for a successful assay include standardizing materials, testing a range of concentrations to provide a dose response curve, applying simple techniques with the possibility of automation, and measuring cell survival with a clear interpretation. 32 Rather than calling for an end to predictive assays, the authors recognized that the need for such an assay has even more appeal as new agents become available. Making rational choices between increasing numbers of regimens is not possible beyond matching a particular histology to a regimen. A method of enhancing responders on a given regimen or excluding nonresponders would be of great benefit to sarcoma patients. These studies concluded that future assays should be studied in the context of a clinical trial, particularly for diseases with a short median survival, when the window for multiple, sequential therapeutic regimens is narrow and when performance status can rapidly decline and thus preclude additional therapies. The importance of a predictive assay would be even more appealing when there is a choice among multiple agents with relatively low percentages of activity, as is the case with STS Despite substantial progress made in recent years for other malignant diagnoses, no clinically validated tests are currently available to predict the efficacy of a given agent for an individual patient with STS other than GIST. Although there is consensus supporting the need to develop and integrate the evaluation of predictive July 2011, Vol. 18, No. 3 Cancer Control 189

3 biomarkers in clinical trials, the practical application of such an approach is still lacking. 33 Assay-Guided Therapy Trials and Retrospective Analyses A search of PubMed literature was conducted that included the following terms and combinations thereof: drug screening assays, antitumor, predictive assay, ex vivo, cancer prediction, individualized medicine, chemosensitive, chemoresistant, in vitro, biomarker, prediction, cancer, prognosis, clinical trial, and sarcoma. Articles were chosen based on review of the abstract and English language. In the following, we present a spectrum of methods and highlight the strengths and limitations of each method. Clinical trials are emphasized, particularly in sarcoma, though these are limited (Table 1) Cell Viability Assays The concept of utilizing tumor cells exposed to chemotherapeutic agents and extrapolating the effects to in vivo disease is as old as modern chemotherapy. Growth inhibition or cell death has been used in previous iterations of assays of sensitivity to conventional chemotherapeutic agents However, due to poor tumor growth under assay conditions, methods that are labor-intensive and time-consuming, and the use of uncertain criteria for defining sensitivity or resistance, these assays have not yet gained wide clinical acceptance. Several methods have been explored using the cell-based platform. The human tumor cloning assay (HTCA) and capillary cloning system (CCS) utilize tumor cell proliferation, measured as the number of colony-forming units in solid media over a period of 2 to 4 weeks. 37,45 While these assays measure tumor cell growth, they remove the tumor microenvironment and the three-dimensionality of an in situ tumor. The techniques are also labor-intensive, taking weeks to complete. They have been studied in a relatively large number of patients and histologic subtypes. One trial reported a 64% rate of successful assay completion. 37 In this trial, a range of malignancies had successfully completed assays, and those treated according to the assay results had an overall response rate of 25%. Patients placed on empiric therapy due to an unsuccessful assay had a response rate of 14%. Of the 470 patients, 14 (3%) had sarcoma. 37 The above assays measure tumor cell growth but do not take into account tumor cells that are alive but not dividing. This important distinction, which can be clinically relevant, is incorporated into the differential staining cytotoxicity (DiSC) assay. This assay uses cells cultured from the primary tumor and then counts live cells after incubation with agents for 4 to 6 days. Modified versions of this test have been used in lung cancer trials in the past with a trend toward improved response-based assay results. Multiple trial designs, including selections of combinations of therapy, have been attempted. 36,46,47 A series that included limited-stage small cell lung cancer patients demonstrated that the biopsy needed for the test Table 1. Selected Studies Evaluating Predictive Assays in Cancer Therapy Study/Assay Method Overview Successful Assay Outcome ATP 34 Multicenter study of metastatic melanoma patients 98% successful; 71% with successful assay received therapy SRCA 35 Primary tumor cells implanted under renal capsule 90% tumor growth in xenograft and assessable after chemotherapeutic therapy Response 36.4% in chemosensitive patients vs 16.1% in chemoresistant patients; overall survival 14.6 mos in chemosensitive patients vs 7.4 mos in chemoresistant patients 11 patients received a tested combination of chemotherapy; 6 of 7 with relapse had corresponding nonresponder status of xenograft DiSC 36 Dye exclusion study, cells treated at reference range, 10% of reference range and 1000% of reference range. Sensitivity < 50% of control in individual drug assay (7 drugs tested); 3-drug combination starting at week 13 from among 13 combinations proven active in small cell lung cancer 44% of assays completed in 12-wk period All patients had a complete or partial response to fi rst 12 wks of standard radiation therapy and chemotherapy; among group of 8 who received assayguided therapy, median survival was 38.5 vs 17.5 mos HTCA 37,38 Cells from primary tumor grown in vitro in the presence of chemotherapeutic assay and colony counts done Multiple trials: 45% to 64% successful 470 patients on single trial Reported higher response rates but nonrandomized ATP 39 Extremity sarcomas were minced, enzymatically digested tumors plated in 96 well plates, and tested against a panel of chemotherapy 100% successful No clinical data available ATP = adenosine 5ʹ-triphosphate, SRCA = subrenal capsule assay, DiSC = differential staining cytotoxicity, HTCA = human tumor clonogenic assay. 190 Cancer Control July 2011, Vol. 18, No. 3

4 did not delay initial chemotherapy and could be used to guide the second 12 weeks of therapy. In addition, in 44% of patients, the median survival was 38.5 months vs 17.5 months in the standard care cohort. While the assay tested seven different agents as monotherapy, the patients were treated with a three-drug combination of assay-active agents that had proven efficacy in this disease type. 36 While the above-described assays may eventually be automated, this technology does not currently exist. Higher-volume analyses using MTT and adenosine 5ʹ-triphosphate (ATP)-based chemiluminescence assays have utilized microplate readers to measure chemical surrogates of cell survival or proliferation. These have been investigated clinically in a variety of tumor types and are readily automated. 48 Whether they accurately correlate to cell survival or number is a matter of ongoing debate. Nevertheless, a multicenter melanoma study demonstrated an overall survival advantage in chemosensitive patients of 14.6 months compared to 7.4 months in chemoresistant patients. 34 In this study, combination cytotoxic chemotherapy was analyzed in the ATP-based assay. The authors concluded that further optimization regarding in vitro drug dosages and combinations tested may need to be explored, though this strategy is currently being tested in the phase III setting through the Dermatologic Cooperative Oncology Group. 34 Importantly, this trial used combinations not routinely used in melanoma and found higher response rates than those reported in historical data. Feasibility of chemosensitivity testing with an ATP-based assay in an exclusive sarcoma population was performed showing that results can be obtained. Unfortunately, no clinical response data was included in the published study. 39 Genome-Wide Microarray Assays Genome-wide microarray assays allow for a snapshot of the transcribed genes and level of transcript in a given tumor. These assays have been used to explore aberrant pathways and to define signatures of tumor subtypes typically within a given histology. Genome-wide microarray assays have been investigated retrospectively due to the cost and time necessary to analyze the large volume of data. Recent reports in bone sarcomas have demonstrated that enriched subsets of genes produce reliable predictive signatures compared to clinical history. This was validated prospectively in 1 case, compared to histologic tumor necrosis in 8 of 8 patients Ongoing studies in sarcoma appear to be underway comparing gene signatures to outcome based on observed abstracts at meetings; however, publications remain limited to biology without clinical outcomes data. 52 With large data sets and numbers of genes upregulated and downregulated in a given experiment, it is often difficult to sort out the relative contributions of the individual genes. Furthermore, the data is not necessarily correlated with protein expression and is unable to identify protein modifications that may be important in the pathogenesis of a given tumor. Inherent biases built into the methodologies also complicate the interpretation of results. 53 Protein Assays Putative markers of chemotherapy resistance including efflux pumps, enhanced DNA repair, mutated target proteins, and variation of metabolizing enzymes have been thoroughly investigated with varying results. Examples of single proteins with prognostic and therapeutic importance can be readily found in cancer literature, with a notable example being hormone receptors in breast cancer. However, markers that affect therapeutic decision making have yet to be developed in non-gist STS. Examples of clinical trials in sarcoma include investigating the efflux pump P-glycoprotein (P-gp) and O 6 - methylguanine-dna-methyltransferase (MGMT). P-gp levels were assessed by immunohistochemistry in an osteosarcoma clinical trial that included P-gp substrates doxorubicin and methotrexate in the treatment of all patients. There was no difference in overall or event-free survival in this study based on P-gp expression. 54 MGMT, which repairs DNA damage due to methylation and is inhibited by O 6 -benzylguanine, varied considerably in the blood of patients with STS, and targeted inhibition of MGMT did not produce any objective responses. 55 Xenograft Assays Patient-derived mouse xenograft models allow for recapitulation of three-dimensional tumor architecture and incorporate aspects of tumor stroma. Two major limitations of animal studies are cost and labor intensity. While sarcoma xenografts have been generated, a comparison of xenograft response to clinical outcome data has not yet been published. One group reported establishment of a series of xenografts that were treated with conventional cytotoxic agents and analyzed for biologic markers of resistance including MDR1, topoisomerase IIα, and glutathione S-transferase. This panel showed low responses to conventional agents and did not have a strong correlation between selected gene markers and chemotherapeutic effect. 56 Xenografts have been used recently in non small cell lung cancer (NSCLC) to explore the chemosensitivity of selected agents or combinations. The subrenal capsule assay uses subcapsular renal transplants of human tumors into murine xenografts treated with various chemotherapies in this disease, with a recently published report establishing a series of xenografts and treating them with three chemotherapeutic combinations. 35 A 90% engraftment rate was achieved, and about half of the tumors were tested against all three chemotherapeutic regimens. The xenografts resembled the primary tumor histologically. Almost a third of the tumors were not sensitive to any of the tested combinations. Perhaps most importantly, a series of 11 patients received one July 2011, Vol. 18, No. 3 Cancer Control 191

5 of the tested combinations (vinorelbine and cisplatin). Seven of these patients recurred, 6 with corresponding nonresponsive xenografts. One xenograft was sensitive to the combination and recurred. Results from this assay took 6 to 8 weeks. The authors concluded that there is a correlation between the assay results and clinical data in the recurrence group and that xenograft-based testing may help identify new active combinations in NSCLC. 35 Summary of Predictive Assay Approaches Individual gene tests, translocation analysis, and protein tests are currently widely used in breast cancer, leukemia, and other malignancies to tailor therapy. For example, FLT3 testing modifies prognosis for acute myelogenous leukemia patients, and therapy with a targeted tyrosine kinase inhibitor molecule or a lower threshold for allogeneic transplantation is being explored in clinical trials at this time. Similarly, the BCR-ABL translocation is used to guide therapy, also with a tyrosine kinase inhibitor in chronic myelogenous leukemia and childhood acute lymphoblastic leukemia. However, no cell-based assays or genome-wide analyses are currently used for the selection of chemotherapy as the accepted standard of care. In non-gist STS, histology and clinical features are often used to guide therapy, but predictive assays are currently not routinely used, if at all. [ 18 F]fluorodeoxy-D-glucose positron emission tomography (FDG-PET) scans are being evaluated in sarcomas to determine the extent of disease and residual disease. One study of 46 patients with STS showed that baseline PET intensity and metabolic response were predictors of relapse-free survival, and a combination of baseline signal intensity and metabolic response was a robust predictor of overall survival. 57 A recent study of high-grade bone sarcomas showed good negative correlation of PET avidity changes after one cycle of chemotherapy with tumor necrosis, with a threshold signal reduction of 60% after one cycle of chemotherapy. 58 However, a sarcoma cohort including intermediate- and high-grade tumors was more complex, with lower sensitivity and specificity. 59 A recent task force report has no recommendation on the routine use of a PET scan during therapy to follow response or to predict response in sarcomas. 60 Many STS clinical trials have compared cytotoxic agents alone or in combination with a spectrum of response rates. While clinical data have led to some insight in matching tumor types that tend to have better response rates, a mechanism-based analysis of tumor tissue treated with chemotherapy is lacking. Doxorubicin has remained the standard of care for 35 years. 6 It is not always clear how cytotoxic agents, which damage DNA or inhibit critical pathways in cells, induce a response. While mechanisms of action of these agents are postulated, biomarkers of efficacy or resistance remain largely unknown. Markers of DNA damage, apoptosis, autophagy, and cell cycle arrest have been identified, though these pathways are complex, interconnected, and often redundant (Table 2). Traditional approaches to the preclinical investigation of novel cancer therapies rely on the use of established human cancer cell lines. These cell lines are maintained in vitro in serum-based growth media, and their responses to experimental cancer therapeutic agents are assessed by studying growth and apoptosis in vitro or in vivo. Prolonged culture of human cancer cells in serum and on tissue culture plastic results in cell lines that may not be Table 2. Potential Molecular Biomarkers of Chemotherapeutic Effect or Resistance Protein Function Pathway PARP Posttranscriptional modifi cation of cellular proteins through poly(adp-ribose)polymerase; this protein is cleaved by caspase-3 during apoptosis Apoptosis RB Tumor suppressor that interacts with E2F family of proteins and halts G 1 /S transition when hypophosphorylated Cell cycle Caspase-3 Member of a family of proenzymes that signal apoptosis though cleavage of other family members Apoptosis Gamma-H2AX Specifi c phosphorylation of a histone in response to double-stranded DNA breaks DNA damage Chk1 and Chk2 Cdc2 Cdc25 Key signal transducer kinases that are activated in response to diverse genotoxic insults to transmit the checkpoint signals from the proximal checkpoint kinases A cell division control protein, also known as cyclin-dependent kinase 1 (Cdk1) or p34/cdk1, plays a key role in the G 2 to M phase transition Family of dual-specifi city phosphatases that positively regulate the cell division cycle by activating cyclin-dependent protein kinases Cell cycle Cell cycle Cell cycle PCNA Proliferating cell nuclear antigen (PCNA) involved in DNA synthesis and postreplication repair Cell proliferation LC3 Microtubule associated protein that interaction between microtubules and the cytoskeleton activated during autophagy Autophagy 192 Cancer Control July 2011, Vol. 18, No. 3

6 representative of the parent tumor. Such differences are of concern in the study of basic cancer biology and are fundamental to our approach in studying this disease. In particular, culture selection in cell lines may disturb the in vitro relationship between the cancer stem cell and its progeny, and it also removes tumor-stromal interactions that are essential to the three-dimensional biology of solid tumors in vivo. To investigate novel therapeutic and diagnostic strategies with greater accuracy, new preclinical strategies such as biomarker assessments are needed to assess anticancer therapies in order to determine if patients are likely to benefit from a given therapy. We propose that future STS studies should involve evaluations of molecular pathways in primary sarcoma tissue. There must be a focus on the biology of the tumor in addition to the histology. Treatment with clinically relevant agents in short-term ex vivo assays and biomarker analysis with corresponding in vivo xenograft or clinical outcomes data would help establish patterns or proteins that may provide information regarding tumor sensitivity or resistance to a given treatment (Figure). Predictive markers can serve as the basis for inclusion or exclusion for clinical trials that can incorporate targeted agents in addition to a cytotoxic therapy. This may eventually lead to combinations of medications that can improve response rates by exploiting signaling or biologic abnormalities in malignant cells. Conclusions The goals of predictive assays or biomarkers are to match efficacious chemotherapy to an individual patient and to exclude chemotherapy that would cause toxicity without benefit. Diseases with a poor prognosis, genetic heterogeneity, many therapeutic options, and a poor response rate to traditional chemotherapy provide the rationale for predictive studies. A quick, reproducible, patient-specific assay that could test many single agents and combinations based on the mechanisms of the therapeutic agents would be ideal. In the setting of conventional cytotoxic agents, biomarkers of biological efficacy are not always known, though molecular markers of DNA damage, cell cycle arrest, apoptosis, autophagy, and necrosis may provide predictive data. Markers of tumor response to cytotoxic chemotherapy have been explored in breast and colon cancer. While the results for colon cancer are controversial and mixed, markers in breast cancer such as topoisomerase IIα and erbb2 show good correlation to response, and ABCB1 (P-gp) and BCL-2 overexpression show good correlation to resistance. 45,61 A clinical trial similar to that shown in the Figure could be conducted to assess the positive and negative predictive value and match patients to chemotherapy to determine if this can improve response rates, progression-free survival, and overall survival. SD, PR, CR First-Line Therapy PD First-Line Therapy Regimen A SD, PR, CR Continue Sarcoma biopsied and treated with front-line and regimen A E ex vivo for 24 hours and molecular markers analyzed PD Ex vivo Assay Results Best Option Regimen B Regimen C Regimen D Regimen E PD Ex vivo Assay Results Second-Best Option Figure. Ex vivo assay schema. This theoretical trial design incorporates an ex vivo assay strategy for metastatic or unresectable STS patients and would allow for prospective assessment of molecular markers, calculation of positive and negative predictive value to fi rst-line therapy, and match chemotherapy in the second line based on ex vivo assay response. While fi rst-line therapy would be anthracycline-based, second-line therapies could consist of cytotoxic agents or combinations with established effi cacy or with potential, utilizing the assay to match the patient to therapy. SD = stable disease, PR = partial response, CR = complete response, PD = progressive disease. July 2011, Vol. 18, No. 3 Cancer Control 193

7 References 1. Borden EC, Baker LH, Bell RS, et al. Soft tissue sarcomas of adults: state of the translational science. Clin Cancer Res. 2003;9(6): Casali PG, Jost L, Sleijfer S, et al. Soft tissue sarcomas: ESMO clinical recommendations for diagnosis, treatment and follow-up. Ann Oncol. 2008;19(suppl 2):ii89-ii Demetri GD, Baker LH, Benjamin RS, et al. Soft tissue sarcoma. J Natl Compr Canc Netw. 2007;5(4): Skubitz KM, D Adamo DR. Sarcoma. Mayo Clin Proc. 2007;82(11): Van Glabbeke M, van Oosterom AT, Oosterhuis JW, et al. Prognostic factors for the outcome of chemotherapy in advanced soft tissue sarcoma: an analysis of 2,185 patients treated with anthracycline-containing fi rst-line regimens a European Organization for Research and Treatment of Cancer Soft Tissue and Bone Sarcoma Group Study. J Clin Oncol. 1999;17(1): Benjamin RS, Wiernik PH, Bachur NR. Adriamycin: a new effective agent in the therapy of disseminated sarcomas. Med Pediatr Oncol. 1975;1(1): Demetri GD, Elias AD. Results of single-agent and combination chemotherapy for advanced soft tissue sarcomas: implications for decision making in the clinic. Hematol Oncol Clin North Am. 1995;9(4): Edmonson JH, Ryan LM, Blum RH, et al. Randomized comparison of doxorubicin alone versus ifosfamide plus doxorubicin or mitomycin, doxorubicin, and cisplatin against advanced soft tissue sarcomas. J Clin Oncol. 1993;11(7): Verma S, Younus J, Stys-Norman D, et al. Ifosfamide-based combination chemotherapy in advanced soft-tissue sarcoma: a practice guideline. Curr Oncol. 2007;14(4): Verma S, Younus J, Stys-Norman D, et al. Meta-analysis of ifosfamidebased combination chemotherapy in advanced soft tissue sarcoma. Cancer Treat Rev. 2008;34(4): Anderson SE, Keohan ML, D Adamo DR, et al. A retrospective analysis of vinorelbine chemotherapy for patients with previously treated softtissue sarcomas. Sarcoma. 2006;2006: Antman KH, Ryan L, Elias A, et al. Response to ifosfamide and mesna: 124 previously treated patients with metastatic or unresectable sarcoma. J Clin Oncol. 1989;7(1): Hensley ML, Maki R, Venkatraman E, et al. Gemcitabine and docetaxel in patients with unresectable leiomyosarcoma: results of a phase II trial. J Clin Oncol. 2002;20(12): Kuttesch JF Jr, Krailo MD, Madden T, et al. Phase II evaluation of intravenous vinorelbine (Navelbine) in recurrent or refractory pediatric malignancies: a Children s Oncology Group study. Pediatr Blood Cancer. 2009;53(4): Leu KM, Ostruszka LJ, Shewach D, et al. Laboratory and clinical evidence of synergistic cytotoxicity of sequential treatment with gemcitabine followed by docetaxel in the treatment of sarcoma. J Clin Oncol. 2004;22(9): Maki RG. Gemcitabine and docetaxel in metastatic sarcoma: past, present, and future. Oncologist. 2007;12(8): Maki RG, Wathen JK, Patel SR, et al. Randomized phase II study of gemcitabine and docetaxel compared with gemcitabine alone in patients with metastatic soft tissue sarcomas: results of sarcoma alliance for research through collaboration study 002 [corrected]. J Clin Oncol. 2007;25(19): Nielsen OS, Judson I, van Hoesel Q, et al. Effect of high-dose ifosfamide in advanced soft tissue sarcomas: a multicentre phase II study of the EORTC Soft Tissue and Bone Sarcoma Group. Eur J Cancer. 2000;36(1): van Oosterom AT, Mouridsen HT, Nielsen OS, et al. Results of randomised studies of the EORTC Soft Tissue and Bone Sarcoma Group (STBSG) with two different ifosfamide regimens in fi rst- and second-line chemotherapy in advanced soft tissue sarcoma patients. Eur J Cancer. 2002;38(18): Karavasilis V, Seddon BM, Ashley S, et al. Signifi cant clinical benefi t of fi rst-line palliative chemotherapy in advanced soft-tissue sarcoma: retrospective analysis and identification of prognostic factors in 488 patients. Cancer. 2008;112(7): Spunt SL, Hill DA, Motosue AM, et al. Clinical features and outcome of initially unresected nonmetastatic pediatric nonrhabdomyosarcoma soft tissue sarcoma. J Clin Oncol. 2002;20(15): Spunt SL, Pappo AS. Childhood nonrhabdomyosarcoma soft tissue sarcomas are not adult-type tumors. J Clin Oncol. 2006;24(12): ; author reply Spunt SL, Poquette CA, Hurt YS, et al. Prognostic factors for children and adolescents with surgically resected nonrhabdomyosarcoma soft tissue sarcoma: an analysis of 121 patients treated at St Jude Children s Research Hospital. J Clin Oncol. 1999;17(12): Spunt SL, Skapek SX, Coffi n CM. Pediatric nonrhabdomyosarcoma soft tissue sarcomas. Oncologist. 2008;13(6): Grosso F, Jones RL, Demetri GD, et al. Effi cacy of trabectedin (ecteinascidin-743) in advanced pretreated myxoid liposarcomas: a retrospective study. Lancet Oncol. 2007;8(7): Fury MG, Antonescu CR, Van Zee KJ, et al. A 14-year retrospective review of angiosarcoma: clinical characteristics, prognostic factors, and treatment outcomes with surgery and chemotherapy. Cancer J. 2005;11(3): Demetri GD, von Mehren M, Blanke CD, et al. Effi cacy and safety of imatinib mesylate in advanced gastrointestinal stromal tumors. N Engl J Med. 2002;347(7): Van Glabbeke M, Verweij J, Judson I, et al. Progression-free rate as the principal end-point for phase II trials in soft-tissue sarcomas. Eur J Cancer. 2002;38(4): Jones RL, Fisher C, Al-Muderis O, et al. Differential sensitivity of liposarcoma subtypes to chemotherapy. Eur J Cancer. 2005;41(18): Samson DJ, Seidenfeld J, Ziegler K, et al. Chemotherapy sensitivity and resistance assays: a systematic review. J Clin Oncol. 2004;22(17): Schrag D, Garewal HS, Burstein HJ, et al. American Society of Clinical Oncology Technology Assessment: chemotherapy sensitivity and resistance assays. J Clin Oncol. 2004;22(17): Schinkothe T, Haeger S, Gabri MR. Practical guidelines for diagnostic use of in vitro chemosensitivity tests. Anticancer Res. 2007;27(3A): Twombly R. Identity crisis: fi nding, defi ning, and integrating biomarkers still a challenge. J Natl Cancer Inst. 2006;98(1): Ugurel S, Schadendorf D, Pfohler C, et al. In vitro drug sensitivity predicts response and survival after individualized sensitivity-directed chemotherapy in metastatic melanoma: a multicenter phase II trial of the Dermatologic Cooperative Oncology Group. Clin Cancer Res. 2006;12(18): Dong X, Guan J, English JC, et al. Patient-derived first generation xenografts of non-small cell lung cancers: promising tools for predicting drug responses for personalized chemotherapy. Clin Cancer Res. 16(5): Cortazar P, Gazdar AF, Woods E, et al. Survival of patients with limitedstage small cell lung cancer treated with individualized chemotherapy selected by in vitro drug sensitivity testing. Clin Cancer Res. 1997;3(5): Von Hoff DD, Clark GM, Stogdill BJ, et al. Prospective clinical trial of a human tumor cloning system. Cancer Res. 1983;43(4): Von Hoff DD, Kronmal R, Salmon SE, et al. A Southwest Oncology Group study on the use of a human tumor cloning assay for predicting response in patients with ovarian cancer. Cancer. 1991;67(1): Lehnhardt M, Muehlberger T, Kuhnen C, et al. Feasibility of chemosensitivity testing in soft tissue sarcomas. World J Surg Oncol. 2005;3(1): Hamburger AW, Salmon SE. Primary bioassay of human tumor stem cells. Science. 1977;197(4302): Kern DH, Weisenthal LM. Highly specifi c prediction of antineoplastic drug resistance with an in vitro assay using suprapharmacologic drug exposures. J Natl Cancer Inst. 1990;82(7): Meitner PA. The fl uorescent cytoprint assay: a new approach to in vitro chemosensitivity testing. Oncology (Williston Park). 1991;5(9):75-81; discussion 81-72, 85, Hirano Y, Ushiyama T, Suzuki K, et al. Clinical application of an in vitro chemosensitivity test, the Histoculture Drug Response Assay, to urological cancers: wide distribution of inhibition rates in bladder cancer and renal cell cancer. Urol Res. 1999;27(6): Sharma S, Neale MH, Di Nicolantonio F, et al. Outcome of ATPbased tumor chemosensitivity assay directed chemotherapy in heavily pretreated recurrent ovarian carcinoma. BMC Cancer. 2003;3: Sekine I, Shimizu C, Nishio K, et al. A literature review of molecular markers predictive of clinical response to cytotoxic chemotherapy in patients with breast cancer. Int J Clin Oncol. 2009;14(2): Gazdar AF, Steinberg SM, Russell EK, et al. Correlation of in vitro drug-sensitivity testing results with response to chemotherapy and survival in extensive-stage small cell lung cancer: a prospective clinical trial. J Natl Cancer Inst. 1990;82(2): Wilbur DW, Camacho ES, Hilliard DA, et al. Chemotherapy of nonsmall cell lung carcinoma guided by an in vitro drug resistance assay measuring total tumour cell kill. Br J Cancer. 1992;65(1): Sargent JM. The use of the MTT assay to study drug resistance in fresh tumour samples. Recent Results Cancer Res. 2003;161: Rabin K, Man TK, Lau CC. Personalized care of pediatric cancer patients. Nestle Nutr Workshop Ser Pediatr Program. 2008;62: ; discussion Man TK, Chintagumpala M, Visvanathan J, et al. Expression profi les of osteosarcoma that can predict response to chemotherapy. Cancer Res. 2005;65(18): Schaefer KL, Eisenacher M, Braun Y, et al. Microarray analysis of Ewing s sarcoma family of tumours reveals characteristic gene expression signatures associated with metastasis and resistance to chemotherapy. Eur J Cancer. 2008;44(5): Baird K, Davis S, Antonescu CR, et al. Gene expression profi l- ing of human sarcomas: insights into sarcoma biology. Cancer Res. 2005;65(20): Kuznetsov VA. Relative avidity, specifi city, and sensitivity of transcription factor-dna binding in genome-scale experiments. Methods Mol 194 Cancer Control July 2011, Vol. 18, No. 3

8 Biol. 2009;563: Schwartz CL, Gorlick R, Teot L, et al. Multiple drug resistance in osteogenic sarcoma: INT0133 from the Children s Oncology Group. J Clin Oncol. 2007;25(15): Ryan CW, Dolan ME, Brockstein BB, et al. A phase II trial of O6-benzylguanine and carmustine in patients with advanced soft tissue sarcoma. Cancer Chemother Pharmacol. 2006;58(5): Boven E, Pinedo HM, van Hattum AH, et al. Characterization of human soft-tissue sarcoma xenografts for use in secondary drug screening. Br J Cancer. 1998;78(12): Schuetze SM, Rubin BP, Vernon C, et al. Use of positron emission tomography in localized extremity soft tissue sarcoma treated with neoadjuvant chemotherapy. Cancer. 2005;103(2): Benz MR, Czernin J, Tap WD, et al. FDG-PET/CT imaging predicts histopathologic treatment responses after neoadjuvant therapy in adult primary bone sarcomas. Sarcoma. 2010;2010: Dimitrakopoulou-Strauss A, Strauss LG, Egerer G, et al. Impact of dynamic 18F-FDG PET on the early prediction of therapy outcome in patients with high-risk soft-tissue sarcomas after neoadjuvant chemotherapy: a feasibility study. J Nucl Med. 51(4): Podoloff DA, Ball DW, Ben-Josef E, et al. NCCN task force: clinical utility of PET in a variety of tumor types. J Natl Compr Canc Netw. 2009;7(suppl 2):S1-S Koopman M, Venderbosch S, Nagtegaal ID, et al. A review on the use of molecular markers of cytotoxic therapy for colorectal cancer, what have we learned? Eur J Cancer. 2009;45(11): July 2011, Vol. 18, No. 3 Cancer Control 195