WO2007013946A2 - Method of predicting responsiveness to chemotherapy and selecting treatments - Google Patents

Abstract

Methods are provided for predicting responsiveness of cancer cells to chemotherapy by measuring the level of phosphorylated Stat or the level of expression of Survivin in a cancer and comparing the level in the cancer cell to the respective level in a control. Also provided are methods of selecting a chemotherapeutic treatment for a subject diagnosed with cancer by measuring the level of phosphorylated Stat or the level of expression of Survivin in a cancer and comparing the level in the cancer cell to the respective level in a control. Kits for performing the methods are also provided. Methods for modulating Survivin-dependent apoptosis in a cancer cell are also disclosed.

Description

METHODS OF PREDICTING RESPONSIVENESS TO CHEMOTHERAPEUTIC AGENTS AND SELECTING TREATMENTS

CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims priority to U.S. Provisional Application No. 60/700,948, filed July 20, 2005 and U.S. Provisional Application No. 60/596,151, filed September 2, 2005. Both provisional applications are incorporated herein by reference.

STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT

[0002] This invention was made with U.S. Government support from the National Institutes of Health Grant Nos. CA 82533, CA 55652, ROl 102726, UOl CA101222. This work was also supported in part by the Chiles Endowment Biomedical Research Program of the Florida Department of Health and by the H. Lee Moffitt Cancer Center & Research Institute. The Government has certain rights in this invention.

BACKGROUND

[0003] Signal transducer and activator of transcription (Stat)-family proteins are latent cytoplasmic transcription factors that convey signals from the cell surface to the nucleus on activation by cytokines and growth factors. See Yu and Jove, Nat Rev 4:97-105 (2004) and Levy and Darnell, Nat Rev MoI Cell Biol 3:651-662 (2002). Engagement of cell surface receptors by polypeptide ligands, such as interleukin-6 (IL-6) or epidermal growth factor, induces tyrosine phosphorylation of Stat proteins by Janus kinase, growth factor receptor tyrosine kinases, and Src family tyrosine kinases. The phosphorylated Stat protein in the activated dimeric form then translocates to the nucleus and regulates expression of genes having Stat-binding sites in their promoters. Under normal physiologic conditions, activation of Stat proteins is rapid, transient and regulates expression of genes that control fundamental biological processes, including cell proliferation, survival, and development. [0004] Numerous studies have detected constitutively active Stat, particularly Statl, Stat3 and Stat5, in diverse human tumor specimens, including myeloma, leukemia, lymphoma, melanoma and carcinomas from prostate, ovary and head and neck. Persistent Stat activity is established as essential for malignant transformation of cultured cells by many oncogenic signaling pathways. For example, the Src, Janus kinase, and epidermal growth factor receptor family tyrosine kinases are frequently activated in breast cancer cells and induce Stat3 activation. Blocking tyrosine kinase pathways with selective pharmacologic inhibitors results in decreased Stat3 activity, growth inhibition, and apoptosis. Persistent activation of Stat3 and Stat5 in tumor cells has been shown to participate in regulating expression of genes involved in controlling cell cycle progression, apoptosis, and angiogenesis. For instance, an oncogenic mutant of Stat3 induces expression of cyclin Dl, Bcl-xL, and c-Myc. [0005] Identification of molecular markers may help guide physicians in the selection of an appropriate chemotherapeutic agent. Identification of molecular markers may aid in the development of target-specific therapies and guide the utilization of such specific chemotherapies.

SUMMARY OF THE INVENTION

[0006] In one aspect, methods of predicting responsiveness of cancer cells to chemotherapy are provided. The level of phosphorylated Stat in a cancer cell is measured and compared to the level of phosphorylated Stat in a control. The level of phosphorylated Stat in the cancer cell as compared to the control is predictive of responsiveness to chemotherapy. Kits for performing the methods are also provided and include an antibody capable of binding phosphorylated Stat.

[0007] In another aspect, methods of predicting responsiveness of a cancer cell to chemotherapy are provided in which the level of expression of Survivin in a cancer cell is measured. The level of expression of Survivin in the cancer cell is then compared to the level of expression of Survivin in a control. The level of expression of Survivin in the cancer cell as compared to the control is predictive of responsiveness to chemotherapy. Kits for performing the methods are also provided and include an antibody capable of binding Survivin or at least two oligonucleotides capable of amplifying a polynucleotide encoding Survivin or an oligonucleotide capable of hybridizing to Survivin mRNA. [0008] In yet another aspect, methods for selecting a chemotherapeutic treatment for a subject diagnosed with cancer are provided in which the level of phosphorylated Stat in a cancer cell from a subject is measured. The level of phosphorylated Stat in the cancer cell is compared to the level of phosphorylated Stat in a control and the comparison is used to select a chemotherapy treatment for the subject which has an expected benefit based on the level phosphorylated Stat in the cancer cell.

[0009] In yet another aspect, methods for treating a subject diagnosed with cancer are provided in which the level of phosphorylated Stat in a cancer cell from the subject is measured. The level of phosphorylated Stat in the cancer cell is compared to the level of phosphorylated Stat in a control and the comparison is used to select a chemotherapy treatment for the subject which has an expected benefit based on the level phosphorylated Stat in the cancer cell. An effective amount of the chemotherapeutic agent is administered to the subject to treat the cancer.

[0010] In a still further aspect, methods of modulating Survivin-dependent apoptosis in a cancer cell are provided in which the cell is contacted with an effective amount of a Stat inhibitor.

BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure IA is a graph representing the decrease in Stat3 mRNA expression after treatment with a Stat3 oligonucleotide. Figure IB is an electrophorectic mobility shift assay demonstrating Stat3 activation is blocked by addition of Stat3 antisense oligonucleotides. [0012] Figure 2A shows light micrographs of normal cells and those treated with Stat3 antisense oligonucleotides. Figure 2B shows light micrographs of an in situ TUNEL assay. Figure 2C is a graph showing the results of the TUNEL assay. Figure 2D is a Western blot analysis detecting PARP cleavage as an indicator of apoptosis. Figure 2E is a graph of viable cells after treatment with Stat3 antisense as compared to control cells. [0013] Figure 3 A is a graph of an RNase protection assay for antiapoptotic genes expressed in a breast cancer cell line after treatment with the Stat3 antisense oligonucleotide. Figure 3B is a Western blot analysis showing expression of Stat3 and Survivin after treatment with Stat3 antisense oligonucleotides. Figure 3 C is a Western blot analysis for PARP cleavage and Survivin expression in breast cancer cells. Figure 3D is a light micrograph of a TUNEL assay.

[0014] Figure 4A is a graph of a luciferase assay to monitor activation of the Survivin promoter in response to the indicated treatments. Figure 4B is a photograph of a chromatin immunoprecipitation assay demonstrating that Stat3 binds to the Survivin promoter region. Figure 4C is a photograph of an EMSA of Stat3 binding activity to the putative Stat3 binding sites in the Survivin promoter region. Figure 4D is a schematic view of the Survivin promoter showing the putative Stat3 binding sites.

[0015] Figure 5 A is a set of photographs of an EMSA for Stat3, and a Western blot for Survivin and Actin in several breast cancer cell lines. Figure 5B is a photograph of an EMSA analysis of Stat3 demonstrating a supershift after IL-6 induction. Figure 5C is a photograph of a Western blot analysis demonstrating induction of phosphorylated Stat3 and increases in Survivin expression after treatment with IL-6.

[0016] Figure 6 is a set of photographs showing immunohistochemical staining for HER2/neu (A and B), pY-Src (C and D) and pY-Stat3 (E and F) on nonneoplastic tissue (A, C, and E) and paired carcinoma (B, D, and F). [0017] Figure 7 is a set of photographs showing immunohistochemical staining for phosphor-Stat3 and Survivin in primary breast tumors and non-neoplastic tissue. [0018] Figure 8 is a graph demonstrating the relationship between response to neoadjuvant chemotherapy treatment and pretreatment levels of phosphorylated Stat3. [0019] Figure 9 A is a graph demonstrating differential sensitivity of non-small cell lung cancer cell lines to gefitinib. Figure 9B is a photograph of a Western blot analysis for phosphorylated and total Stat3. Figure 9C is a photograph of a Western blot analysis for phosphorylated Stat3 and total Stat3 after treatment with increasing amounts of gefitinib. Figure 9D is a photograph of a Western blot analysis for Stat3, cleaved PARP and Actin after treatment with Stat3 antisense oligonucleotides (AS) or a mismatch control (MM). Figure 9E is a photograph of a Western blot analysis for the indicated apoptosis-related markers after treatment with Stat3 antisense oligonucleotides (AS) or a mismatch control (MM).

DETAILED DESCRIPTION

[0020] Numerous studies have demonstrated that persistent activation of Stats, particularly Statl, Stat3 and Stat5a and Stat5b, occurs in a wide variety of tumors, including myeloma, leukemia, lymphoma, melanoma, and carcinoma from prostrate, ovary, breast, lung, pancreatic and head and neck. Stats are believed to contribute to oncogenesis by several mechanisms including inhibition of apoptosis, enhancement of cell proliferation, induction of angiogenesis, and suppression of immune responses. The Examples below describe a relationship between activated (phosphorylated) Stat3 and tumor cell responsiveness to various chemotherapeutic agents.

[0021] One of skill in the art will appreciate that different cancer cells and different types of cancer respond differently to different types of chemotherapeutic agents. Even within the same type of cancer, such as breast cancer, some cancer cells are more responsive to certain types of chemotherapy than other cancer cells. Currently, physicians and other primary care providers rely on trial and error to determine which chemotherapeutic agents are effective against a particular cancer. Often, cancer cells are resistant to the chemotherapeutic agent and treatment fails to eliminate the cancer. Thus, a diagnostic test capable of predicting responsiveness to a chemotherapeutic agent would be valuable. [0022] The Examples demonstrate that activated Stat3 induces expression of genes involved in blocking apoptosis, such as Survivin, a member of the inhibitor of apoptosis protein (IAP) family, and that apoptosis is induced if Stat3 is blocked. See Li and Altieri, Biochem J 344:305-311 (1999), which is incorporated herein by reference. Furthermore, the levels of activated Stat3 were found to be correlated to the response of breast cancer tumors from women with stage III breast cancer to neoadjuvant chemotherapy with doxorubicin and docetaxel. Tumors having lower levels of activated Stat3 were shown to be more likely to regress in response to chemotherapy than are tumors with high levels of activated Stat3. In non-small cell lung cancer, cells having higher levels of activated Stat3 were found to be more sensitive to treatment with gefitinib, an EGFR tyrosine kinase inhibitor. Taken together the Examples demonstrate a method of predicting responsiveness of cancer cells to a chemotherapeutic agent.

[0023] The methods involve measuring the level of phosphorylated Stat in a cancer cell and comparing the level of phosphorylated Stat in the cancer cell to the level of phosphorylated Stat in a control. The level of phosphorylated Stat in the cancer cell as compared to the control is predictive of responsiveness to a chemotherapeutic agent. The responsiveness of the cancer cell to chemotherapy may be decreased or increased depending on the chemotherapeutic agent. The relationship between the level of phosphorylated Stat and the responsiveness of the cancer cell to a particular chemotherapeutic agent can be determined by one of skill in the art. For example, an increased level of phosphorylated Stat is predictive of a decreased level of responsiveness to conventional chemotherapies, such as doxorubicin and docetaxel, or other chemotherapeutic agents that induce apoptosis and indicates treatment with a tyrosine kinase inhibitor.

[0024] The methods also include measuring the level of expression of Survivin in a cancer cell and comparing the level of expression of Survivin in the cancer cell to the level of expression of Survivin in a control. The level of expression of Survivin in the cancer cell as compared to the control is predictive of responsiveness to chemotherapy. The responsiveness of the cancer cell to chemotherapy may be decreased or increased depending on the chemotherapeutic agent. The relationship between the level of phosphorylated Stat3 and the responsiveness of the cancer cell to a particular chemotherapeutic agent can be determined by one of skill in the art. For example, an increased level of expression of Survivin is predictive of a decreased level of responsiveness to conventional chemotherapies such as doxorubicin and docetaxel and may indicate treatment with a tyrosine kinase inhibitor. [0025] A cancer cell is responsive to a chemotherapeutic agent if the chemotherapeutic agent induces apoptosis, decreases cell proliferation, or induces an immune response against the cancer cell. Responsiveness of a cancer cell to a chemotherapeutic agent may also be measured as a reduction in tumor size, or inhibition of angiogenesis. Responsiveness to a chemotherapeutic agent may be measured after the therapy is completed by methods known to those of skill in the art including, but not limited to, palpation, imagery and surgical removal. In the Examples, the residual tumor was surgically removed after chemotherapy in individuals with stage III breast cancer and the size of the tumor was compared to the size of the tumor by palpation prior to chemotherapy.

[0026] Responsiveness to the chemotherapeutic agent may be either an in vitro response of a cancer cell or an in vivo clinical response of a cancer in a subject. Cancer includes, but is not limited to, tumors, cancer cells and metastases. The responsiveness of a cancer cell to treatment with a chemotherapeutic agent may be assessed in a variety of ways known to those of skill in the art, including, but not limited to, proliferation assays and apoptosis assays. For example, proliferation assays, include but are not limited to ³H-thymidine incorporation, trypan blue exclusion assays and MTT assays. Apoptosis assays include, but are not limited to, TUNEL, PARP cleavage, and DNA fragmentation. In vivo assays for responsiveness are also known to those of skill in the art and include, but are not limited to, assays for assessing the size and metastasis of a cancerous tumor.

[0027] Clinical response of a tumor to treatment with a chemotherapeutic agent may be assessed by measuring the size of the tumor. One of ordinary skill in the art would understand how to determine the size of a tumor. For example the size may be determined by palpation or imaging, such as a CT scan or an MRI. A complete clinical response is defined as disappearance of all measurable tumor by clinical exam, such as by palpation or imaging. A partial clinical response is defined as a reduction by at least 50% of the size of the tumor. Progression of a tumor is defined as an increase in the size of the tumor or the appearance of any new lesions. Stable disease is defined as a response that did not meet the partial response or progression criteria.

[0028] Pathologic response is determined at the time of surgery, and a response is classified as a complete pathologic response or partial pathologic response based on the residual tumor size after chemotherapy. A complete pathologic response is defined as substantially no evidence of tumor in or around the original site. In the Examples, pathologic response was determined in the breast and separately in the axillary lymph nodes in those patients who were known to have documented positive lymph nodes by palpation, fine needle aspiration, or sentinel lymph node mapping prior to chemotherapy. A partial pathologic response is defined as evidence of tumor, but of a smaller size than the tumor prior to chemotherapy. [0029] The methods in accordance with the present invention are useful in predicting responsiveness of cancers to a particular chemotherapeutic agent based on a statistical analysis as described in the Examples. The Student's t test was used if the data followed a normal distribution and the Wilcoxon Mann- Whitney test was used if the normality assumption was not met. AU tests were two-sided and declared significant at the 5% level. Cancers having high levels of phosphorylated Stat or high levels of Survivin relative to noncancerous control cells are less likely to respond to some chemotherapy agents such as the taxane derivatives or anti-cancer agents such as doxorubicin, but are more likely to respond to chemotherapeutic agents which are inhibitors of tyrosine phosphorylation or other signaling pathways such as Stat inhibitors or EGFR tyrosine kinase inhibitors. For example, Figure 8 in Example 7 demonstrates that individuals with cancer cells having high levels of phosphorylated Stat3 are less likely to mount a complete pathologic response to the breast cancer after treatment with doxorubicin and docetaxel than individuals whose cancer cells have lower levels of phosporylated Stat3. On the other hand, Figure 9 A and 9B in Example 10 demonstrate that cancer cells having higher levels of phosphorylated Stat3, such as H3255, were more susceptible to treatment with gefitinib than cell lines having relatively low levels of phosphorylated Stat3, such as H1299.

[0030] The methods in accordance with the present invention may be used to predict responsiveness of a cancer to a chemotherapeutic agent. The subject diagnosed with cancer may be human or a non-human mammal. The methods allow prediction of responsiveness of a number of different cancers including, but not limited to breast cancer, lung cancer, ovarian cancer, head and neck cancer, melanoma, lymphoma, leukemia, multiple myeloma, prostrate cancer, gastric cancer, colon cancer and pancreatic cancer.

[0031] The methods in accordance with the present invention may be used to predict responsiveness of cancers to various chemotherapeutic regimens. In the Examples, the methods are used to enable prediction of responsiveness of breast cancer to neoadjuvant therapy with doxorubicin and docetaxel and non-small cell lung cancer to EGFR tyrosine kinase inhibitor therapy. Neoadjuvant therapy refers to the use of chemotherapy or hormonal therapy as the initial treatment of a primary malignant tumor followed by surgery to remove the tumor.

[0032] The methods may be used to predict responsiveness to chemotherapeutic agents delivered by any means known to those of skill in the art. The methods can also be used to predict responsiveness of cancers to other chemotherapeutic agents including, but not limited to, Stat inhibitors and Survivin inhibitors. For example, Stat3 inhibitors include, but are not limited to, RNAi directed against Stat3 mRNA and those identified in the following publications: Turkson et al., J Biol Chem 276:45443-455 (2001); Turkson et al, MoI Cancer Ther 3:261-69 (2004); Sun et al., Oncogene 24:3236-3245 (2005); and Turkson et al., MoI Cancer Ther (in revision), which are incorporated herein by reference. [0033] Chemotherapeutic agents include, but are not limited to, alkylating agents such as busulfan, cisplatin, carboplatin, chlorambucil, cyclophosphamide, ifosfamide, dacarbazine (DTIC), mechlorethamine (nitrogen mustard), melphalan, and temozolomide; nitrosoureas such as carmustine (BCNU) and lomustine (CCNU); antimetabolites such as 5-fluorouracil, capecitabine, 6-mercaptopurine, methotrexate, gemcitabine, cytarabine (ara-C), fludarabine, and pemetrexed; anthracyclines such as daunorubicin, doxorubicin (Adriamycin), epirubicin, idarubicin, and mitoxantrone; topoisomerase I or II inhibitors such as topotecan and irinotecan, etoposide (VP- 16) and teniposide; mitotic inhibitors such as the taxanes (paclitaxel, docetaxel) and the vinca alkaloids (vinblastine, vincristine, and vinorelbine); L- asparaginase, dactinomycin, thalidomide, tretinoin, gefitinib, and erlotinib. [0034] The level of phosphorylated Stat may be measured in a variety of ways including, but not limited to, immunohistochemistry, electrophorectic mobility shift assay, Western blot, and enzyme-linked immunosorbent assay (ELISA). Each of these assays is capable of differentiating between phosphorylated Stat and its unactivated, non-phosphorylated form. One of skill in the art will appreciate that for diagnostic purposes, a measurement system adaptable for high throughput analysis such as ELISA, or a spectroscopic or enzymatic assay would be suitable.

[0035] The level of expression of Survivin may be measured at either the protein or RNA level by any method known in the art including, but not limited to, immunohistochemistry, Western blot, ELISA, microarray, rt-PCR and Northern blot. One of skill in the art will appreciate that for diagnostic purposes a measurement system adaptable for high throughput analysis such as ELISA, or a spectroscopic or enzymatic assay would be suitable. [0036] The level of phosphorylated Stat and the level of expression of Survivin may be measured in a cancer cell or a tumor from a subject diagnosed with cancer and in a control. The control may be non-cancerous cells or non-cancerous tissues or cancer cells or cancerous tissues with known responsiveness to chemotherapy. In another embodiment, the control may be a cell line, such as H460, H1299, H549, H358, or MDA-MB-361 cells, having a reference level of phosphorylated Stat or a reference level of expression of Survivin. The control may also be a reference chart for comparison of levels of phosphorylated Stat and expression of Survivin.

[0037] In one embodiment, the control may be a non-cancerous control cell obtained from the same subject as the cancer cell or obtained from a different subject. The control may be obtained by any method known to those of skill in the art including, but not limited to, biopsy, fine needle aspiration, and surgical resection or removal.

[0038] In another embodiment, the control is a set of samples containing known amounts of phosphorylated Stat or a known level of expression of Survivin. The control samples, containing various levels of phosphorylated Stat or levels of expression of Survivin, have known responsiveness to chemotherapy such that the control samples form a curve of responsiveness as it relates to the level of phosphorylated Stat or the level of expression of Survivin. Such a set of control samples provides adequate relative information that a primary care provider can compare the level of phosphorylated Stat or the level of expression of Survivin in a cancer from a subject to the respective level of phosphorylated Stat or level of expression of Survivin in the control samples to make an informed prediction of the responsiveness of the cancer to various types of chemotherapy.

[0039] Methods for selecting a chemotherapeutic agent for a subject diagnosed with cancer are also provided. The methods involve measuring the level of phosphorylated Stat or the level of expression of Survivin in a cancer cell obtained from a subject and comparing the level of phosphorylated Stat or the level of expression of Survivin in the cancer cell to the level of phosphorylated Stat or the level of expression of Survivin, respectively, in a control. This information is then used to determine the chemotherapeutic agent for the subject. Finally, an effective amount of the chemotherapeutic agent is administered to the subject to treat the cancer.

[0040] If a cancer is predicted to be less responsive or more responsive to a particular chemotherapeutic agent by the methods described above, then the chemotherapy treatment provided to the subject can be tailored such that the subject is given the chemotherapy treatment to which the cancer is likely to respond. In the Examples, non-small cell lung cancer cells having increased levels of phosphorylated Stat3 were demonstrated to have increased sensitivity to EGFR tyrosine kinase inhibitors. Thus subjects with non-small cell lung cancer which demonstrate increased levels of phosphorylated Stat3 should be treated with an effective amount of an EGFR tyrosine kinase inhibitor. In patients with breast cancer, cancer cells having an increased level of phosphorylated Stat3 and increased expression of Survivin were less likely to respond to neoadjuvant chemotherapy with doxorubicin and docetaxel. Thus, a subject with breast cancer having an increased level of phosphorylated Stat3 or of Survivin should be treated with an alternative chemotherapy, such as Src inhibitors or Stat3 inhibitors. An increased level of phosphorylated Stat3 in a cancer cell as compared to a control is indicative of sensitivity of the cancer to EGFR tyrosine kinase inhibitors, Stat3 inhibitors and Src kinase inhibitors.

[0041] Treatment or treating a cancer includes, but is not limited to, reduction in cancer growth or tumor burden, enhancement of an anti-cancer immune response, induction of apoptosis of cancer cells, inhibition of angiogenesis, enhancement of cancer cell apoptosis, and inhibition of metastases. Administration of an effective amount of a chemotherapeutic agent to a subject may be carried out by any means known in the art including, but not limited to intraperitoneal, intravenous, intramuscular, subcutaneous, transcutaneous, oral, nasopharyngeal or transmucosal absorption. The specific amount or dosage administered in any given case will be adjusted in accordance with the specific cancer being treated, the condition, including the age and weight, of the subject, and other relevant medical factors known to those of skill in the art.

[0042] Dosages for a particular subject can be determined using conventional considerations known to those of skill in the art, including but not limited to a subject's age, body weight, general state of health, diet, the timing and mode of administration, the rate of excretion, other medicaments used in combination and the severity and invasiveness of the cancer. For example, between 60 and 90 milligrams per square meter of doxorubicin may be administered via a single intravenous (IV) injection every 21 days. Suitably, 75 to 85 mg/m² doxorubicin may be administered. Alternately, between 20 and 30 milligrams of doxorubicin per square meter per day may be given via IV for three days every three to four weeks. Alternately, 20 milligrams of doxorubicin per square meter may be given via IV weekly. The dose of doxorubicin used depends upon which regimen for cancer therapy is being followed. Docetaxel maybe administered intravenously, in a dose that ranges from 60-100 mg/m ², over one hour, once every three weeks. Suitably, docetaxel may be administered in a dose of 70-90mg/m², or from 70-80 mg/m². Gefitinib may be taken orally, once daily at 250mg/day for an adult.

[0043] Methods are also provided for modulation of Survivin-dependent apoptosis in a cancer cell. The cancer cell is contacted with an effective amount of Stat3 inhibitor to induce apoptosis. Stat3 inhibitors include, but are not limited to, those identified by the methods described above and those identified in the following publications: Turkson et al., J Biol Chem 276:45443-455 (2001); Turkson et al., MoI Cancer Ther 3:261-69 (2004); Sun et al., Oncogene 24:3236-3245 (2005); and Turkson et al., MoI Cancer Ther (in revision) which are incorporated herein by reference. As demonstrated in the Examples below, activated Stat3 induces expression of Survivin and Survivin in turn blocks apoptosis. Therefore, blocking or reversing Stat3 activation and phosphorylation results in decreased expression of Survivin and allows induction of apoptosis in some cancer cells. The following examples are provided to assist in further understanding of the invention. The particular materials and methods employed are considered to be illustrative of the invention and are not meant to be limiting on the scope of the claims.

EXAMPLES

Example 1 : Stat3 antisense oligonucleotides block expression and activation of Stat3. [0044] The human breast cancer cell line MDA-MB-435s, harbors activated Stat3. The cells were grown in DMEM supplemented with 10% fetal bovine serum, 100 units/mL penicillin, and 100 μg/mL streptomycin. The cells were transfected with Stat3 antisense or mismatch oligonucleotides using Lipofectamine-Plus as described by the supplier (Life Technologies, Grand Island, NY). The Stat3 antisense (5¹GCTCC AGC ATCTGCTGCTTC- 3') (SEQ ID NO: 1) and control mismatch (5'GCTCCAATACCCGTTGCTTC-3^r) (SEQ E) NO: 2) oligonuceotides were synthesized using phosphorothioate chemistry. To increase stability, oligonucleotides were synthesized with 2'-O-methoxyethyl modification of the five or three underlined terminal nucleotides. The final concentration for Stat3 antisense and control mismatch oligonucleotides was 250 nmol/L. Twenty-four and 36 hours after transfection, the mRNA levels and DNA-binding activities of Stat3 were measured by RNase protection and electrophoretic mobility shift (EMSA) assays, respectively. [0045] RNase protection assay. Total RNA was isolated from MDA-MB-435s cells using the RNAeasy mini kit (Qiagen, Valencia, CA). RNase protection assays were carried out with the Riboquant hStress-1 template set containing Bcl-xL and McI-I probes or custom-made multiprobe templates containing Stat3 probes (BD PharMingen, San Diego, CA). Briefly, the multiprobe templates were synthesized by in vitro transcription with incorporation of [³²P]dUTP and purified on Quick Spin RNA columns (Roche Applied Science, Indianapolis, FN). Labeled probe (1 x 10⁶ cpm) was hybridized with 10 μg of total RNA through a temperature gradient of 90°C to 56°C over a 16-hour period. Unprotected probe was removed by RNase digestion at 30°C for 1 hour followed by separation of protected RNA fragments on a 5% polyacrylamide-urea gel and detection using autoradiography.

[0046] Nuclear extract preparation and EMSA. Nuclear extracts were prepared as previously described by Yu et al., Science 269:81-83 (1995), which is incorporated herein by reference, by high-salt extraction into 30 to 70 μL buffer [20 mmol/L HEPES (pH 7.9), 420 mmol/L NaCl, 1 mmol/L EDTA, 20% glycerol, 20 mmol/L NaF, 1 mmol/L Na₃VO₄, 1 mmol/L Na₄P₂O₇, 1 mmol/L DTT, 0.5 mmol/L phenylmethylsulfonyl fluoride, 0.1 μmol/L aprotinin, 1 μmol/L leupeptin, and 1 μmol/L antipain].- For EMSA, 5 μg of total nuclear protein were used for each lane. EMSA was done using a ³²P-labeled oligonucleotide probe containing a high-affinity cw-inducible element (hSIE, m67 variant) derived from the c-fos gene promoter (sense strand 5ΑGCTTCATTTCCCGTAAATCCCTA-3') (SEQ E) NO: 3) that binds activated Stat3 proteins. Following incubation of radiolabeled probes with nuclear extracts, protein-DNA complexes were resolved by nondenaturing PAGE and detected by autoradiography. Stat3 protein was supershifted in the EMSA by preincubation with Stat3 antibody (C-20X, Santa Cruz Biotechnology, Santa Cruz, CA). [0047] Figure IA shows that Stat3 antisense diminished Stat3 mRNA expression compared with the mismatch oligonucleotides. The decrease in mRNA expression was accompanied by a significant decrease in Stat3 DNA-binding activity (Fig. IB). Example 2: Direct blocking of Stat3 induces apoptosis in breast cancer cells. [0048] To assess whether apoptosis was occurring in the cells after blocking of Stat3 the cells were examined microscopically and a TUNEL assay and poly-(ADP)-ribose polymerase (PARP) cleavage assay were performed.

[0049] In situ terminal deoxyribonucleotidyl transferase-mediated dUTP nick end labeling and cellular proliferation assays. MDA-MB-435s and MDA-MB-231 cells were transfected with antisense or control mismatch oligonucleotides. After 48 hours, cells were labeled for apoptotic DNA strand breaks by terminal deoxyribonucleotidyl transferase- mediated dUTP nick end labeling (TUNEL) reaction using an in situ cell death detection assay (Roche Applied Science, Indianapolis, IN) according to the instructions of the supplier. TUNEL-positive nuclei were counted and the apoptotic index was expressed as the number of apoptotic cells in one microscopic field. To determine cellular viability, cells were harvested by trypsinization and counted by trypan blue exclusion assay at 24 and 48 hours after transfection. All experiments were done in triplicate.

[0050] Western blot analyses. Cells were lysed in a buffer containing 10 mmol/L Tris- HCl (pH 8.0), 150 mmol/L NaCl, 1% NP40, 0.5% sodium deoxycholate, 0.1% SDS, 1 mmol/L EDTA (pH 8.0), 2 mmol/L phenylmethylsulfonyl fluoride, 2 μg/mL aprotinin, 2 μg/mL leupeptin, and 1 mmol/L Na₃VO₄. For Western blot analyses, 30 μg of total extracted proteins were applied per lane before SDS-PAGE. Following transfer to nitrocellulose membranes, protein expression levels were detected using polyclonal anti-poly-(ADP-ribose)

polymerase antibodies (Cell Signaling Technology, Beverly, MA). The expression of β-actin

(Sigma- Aldrich, St. Louis, MO) was used as a normalization control for protein loading. [0051] Incubation of cells with Stat3 antisense oligonucleotides for up to 48 hours resulted in a marked increase in vacuolated cells and cellular debris (Fig. 2A), indicative of apoptotic cell death. The occurrence of apoptosis was confirmed by in situ TUNEL assay (Fig. 2B and C) and by cleavage of poly-(ADP-ribose) polymerase at 48 hours after treatment with antisense oligonucleotides (Fig. 2D). Figure 2E shows that treatment with Stat3 antisense oligonucleotide also induced significant growth inhibition as measured by trypan blue exclusion after 24 or 48 hours of incubation. Both apoptosis and inhibition of cellular proliferation correlated with blockade of Stat3 expression and activation (See, e.g., Fig. 1 and Gritsko et al., Clin. Cancer Res. 12:11-19 (2006) which is incorporated herein by reference).

Example 3: Inhibition of Stat3 decreases expression of Survivin in breast cancer cells. [0052] Microarray gene expression profiling analyses were completed to assess the gene expression changes associated with blockade of Stat3 activity. Five micrograms of total RNA collected from MDA-MB-435s cells treated with antisense or control mismatch oligonucleotides for 24 hours served as the mRNA sources for microarray analysis. The poly(A) mRNA was specifically converted to cDNA and then amplified and labeled with biotin following the procedure initially described by Van Gelder et al., Proc. Natl. Acad. Sci. USA 87: 1663-67 (1990) which is incorporated herein by reference. Hybridization with the biotin-labeled DNA, staining, and scanning of the microarray chips followed the prescribed procedure outlined in the Affymetix technical manual.

[0053] The oligonucleotide probe arrays were Human Genome Ul 33 A chips

(Affymetrix, Santa Clara, CA). Scanned output files were visually inspected for hybridization artifacts and then analyzed using Affymetrix Microarray MAS 5.0 software. The MAS 5.0 software identifies the increased and decreased genes between any two samples with a statistical algorithm that assesses the behavior of oligonucleotide probe sets designed to detect the same gene. Probe sets that yielded a change at P < 0.0045 were identified as changed (increased or decreased). In addition, the data were processed using robust multiarray analysis as described by Bolstad et al., Bioinformatics 19:185-93 (2003) which is incorporated herein by reference in its entirety. Genes that were significantly changed in their expression were identified. Empirical estimates of the null distribution were determined using permutation analysis, thereby controlling the number of false positives. The significance analysis of microarrays (as described by Tusher et al., Proc Natl Acad Sci USA 98:5116-21 (2001), which is incorporated herein by reference) implements this approach to address the multiple testing problem and was also applied to the data analysis. Genes were considered changed if consistent behavior (increase or decrease) was observed in each of three replicate experiments based on analyses of data by the multiple methods described above.

[0054] The biological functions of the genes identified were diverse, but included several apoptosis-related genes. Microarray analysis did not reveal consistent decreases in the expression of any Bcl-2 family proteins in antisense oligonucleotide-treated breast cancer cells (See Table 1). In contrast, expression of Survivin, which is a member of the IAP family of antiapoptotic genes, was found to be diminished by microarray analysis as confirmed below by independent molecular approaches. The microarray analysis revealed no change in other IAP family members including X-linked LAP, cellular IAP-I, and cellular IAP-2 (Table

1).

Table 1.

[0055] The microarray data were validated using an RNase protection assay using a Survivin probe (BD PharMingen, San Diego, CA) as described above. Results showed a decrease in mRNA expression of Survivin, but not of Bcl-xL or McI-I, in antisense oligonucleotide-treated breast cancer cells (Fig. 3A). The correlation between Stat3 and Survivin protein expression was further confirmed by Western blot analysis. Protein expression levels were detected using polyclonal anti-Stat3 (Santa Cruz Biotechnology, Santa Cruz, CA) and polyclonal anti-Survivin (Alpha Diagnostics International, San Antonio, TX) antibodies. The data demonstrated decreased Survivin expression in breast cancer cells that are treated with Stat3 antisense oligonucleotide (Fig. 3B).

[0056] Activated Stat3 signaling promoted the survival of breast tumor cells, through an unknown mechanism. Because activated Stat3 was shown to induce expression of Survivin and Survivin is a member of the IAP family, the ability of Survivin expression to protect breast cancer cells from apoptosis was investigated. Antisense oligonucleotides directed against Survivin (5'-CCCAGCCTTCCAGCTCCTTG-S'^') (SEQ ID NO: 4) were synthesized using phosphorothioate chemistry as described above and were used at a final concentration of 300 nmol/L. MDA-MB-435s and MDA-MB-231 cells were transfected with antisense oligonucleotides directed against Survivin as described above. PoIy-(ADP -ribose) polymerase cleavage (Fig. 3C) and in situ TUNEL staining (Fig. 3D) were evident following inhibition of Survivin expression, indicative of apoptosis. Thus, expression of the antiapoptotic protein Survivin is associated with constitutive Stat3 activity and survival in breast cancer cells. Example 4: Stat3 directly binds to and regulates the Survivin promoter. [0057] To determine whether Stat3 regulates Survivin promoter activity, transient transfection studies with a luciferase reporter gene driven by the human Survivin promoter were completed. The reporter pGL2-Survivin encodes the Survivin gene promoter driving expression of firefly luciferase in pGL2 (Promega, Madison, WI). Expression vectors for v- Src (pMvSrc) and Stat3 (pVRStat3) have been previously described by Turkson et al., MoI Cell Biol. 18:2545-52 (1998), which is incorporated herein by reference. Cytosolic extract preparation and luciferase assays were done as previously described by Turkson et al., MoI Cell Biol 18:2545-52 (1998). Briefly, cells were lysed in 0.1 mL of low-salt HEPES buffer [10 mmol/L HEPES (pH 7.8), 10 mmol/L KCl, 0.1 mmol/L EGTA, 0.1 mmol/L, EDTA, 1 mmol/L phenylmethylsulfonyl fluoride, 1 mmol/L DTT, and 20 μL of 10% NP40]. After centrifugation (10,000 x g, 1 minute, 4⁰C), cytosolic supernatant was used for luciferase assays as described by the vendor (Promega). Experiments were done in triplicate and the average values were determined. To control for transfection efficiency, firefly luciferase values were normalized to the values for β-galactosidase.

[0058] Cotransfection of the Survivin reporter construct with a v-Src vector that activates endogenous cellular Stat3 induced expression of the Survivin reporter by 2-fold (Fig. 4A). Moreover, cotransfection with both v-Src and full-length Stat3 vectors further induced the Survivin reporter expression up to nearly 5-fold. By contrast, ectopic expression of the

dominant-negative Stat3 variant, Stat3β, decreased basal levels of Survivin reporter

expression by 50% (Fig. 4A).

[0059] A search for potential Stat3-binding sites within the Survivin promoter region revealed five candidates with the consensus sequences TT(N₄)AA (SEQ ID NO: 5) and TT(N₅)AA (SEQ ID NO: 6). The Survivin promoter was described by Li and Altieri, Biochem J 344:305-11 (1999) and the STAT DNA binding sequence was described by Seidel et al., Proc Natl Acad Sci USA 92:3041-45 (1995), both of which are incorporated herein by reference. To determine whether Stat3 could bind the Survivin promoter under physiologic conditions in intact cells, chromatin immunoprecipitation assays were performed using three sets of primers that cover the five candidate Stat3-binding sites. Chromatin immunoprecipitation assays were done as previously described by Wells et al., MoI Cell Biol 20:5797-807 (2000), which is incorporated herein by reference. Briefly, asynchrononously growing HEK-293 cells were incubated with formaldehyde to cross-link protein-DNA complexes. The cross-linked chromatin was then extracted, diluted with lysis buffer, and sheared by sonication. After preclearing with 1 :2 mix of protein A/protein G-agarose beads (Life Technologies, Grand Island, NY), the chromatin was divided into equal samples for immunoprecipitation with either anti-Stat3 or anti-immunoglobulin G (negative control) polyclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA). The immunoprecipitates were pelleted by centrifugation and incubated at 56°C to reverse the protein-DNA cross- linking. The DNA was extracted from the eluate by the phenol/chloroform method and then precipitated by ethanol. Purified DNA was subjected to PCR with primers specific for a region (-1,231 to -1,009) in the human Survivin promoter spanning three putative Stat3- binding sites. The sequences of the PCR primers used are as follows: Survivin forward primer, 5^I-CAGTGAGCTGAGATCATGCC-3^I (SEQ ID NO: 7); Survivin reverse primer, 5'- TATTAGCCCTCCAGCCCCCAC-3' (SEQ ID NO: 8).

[0060] Primers to the region of -1,231 to -1,009 upstream from the ATG translation initiation site yielded Survivin promoter DNA in chromatin immunoprecipitated with an anti- Stat3 antibody (Fig. 4B). This region contains three potential Stat3 binding sites (Fig. 4D). By contrast, primers to the region of -358/-148 and -938/-759 in the Survivin promoter did not detect promoter DNA in the anti-Stat3 immunoprecipitates. [0061] EMSA was done to determine binding of Stat3 to the same Survivin promoter region in vitro. The oligonucleotides containing the putative Stat3-binding sites in the Survivin promoter used in EMSA are as follows (sense strand): (-1,184) site #1, 5'- TGGAGACTCAGTTTCAAATAAATAAATAAAC-S' (SEQ ID NO: 9); (-1,143) site #2, 5'- TGAGTTACTGTATTAAAGAATGGGGGCGGG-S' (SEQ ID NO: 10); and (-1,105) site #3, 5'-TGTGGGGAGAGGTTGC AAAAATAAATAAAT-3' (SEQ TD NO: 11) (the bolded and underlined oligonucleotides, TG, are sequences added to the 5' end to create overhangs for radiolabeling by Klenow reaction and are not part of the Survivin promoter). Competition analysis to determine specificity of Stat3 binding to the Survivin promoter was done by preincubating unlabeled hSIE probe with radiolabeled Survivin probe in the EMSA. Following incubation of radiolabeled probes with nuclear extracts, protein-DNA complexes were resolved by nondenaturing PAGE and detected by autoradiography. [0062] Results showed that endogenous activated Stat3 protein, present in nuclear extracts of MDA-MB-435s breast cancer cells, bound to the Survivin promoter fragments - 1,174/-1,166 (site #1) and -l,095/-l,087 (site #3) but not to fragments -1,133/-1,126 (site #2), -851/-844 (site #4), and -264A256 (site #5; Fig. 4C and data not shown). Specificity of Stat3 binding to the Survivin promoter fragments #1 and #3 was shown by competition analysis with unlabeled hSIE probe (Fig. 4C and data not shown). Both site #1 and site #3 are located within the -1,231 to -1,009 region that was detected in the chromatin immunoprecipitation assays above with anti-Stat3 antibody, suggesting it is this region of the Survivin promoter that accounts for Stat3 binding. Taken together, these data provide evidence that Stat3 directly binds the Survivin promoter and induces its expression.

Example 5: Stat3 activation correlates with expression of Survivin in breast cancer cells. [0063] To investigate whether Stat3 activation correlates with Survivin up-regulation, Western blot analyses for Survivin were completed in a panel of human breast cancer cells harboring constitutively active Stat3 and in normal breast epithelial cells. The Western blots were completed as described above and were probed with polyclonal anti-Survivin antibodies

(Alpha Diagnostics International, San Antonio, TX). The expression of β-actin (Sigma-

Aldrich, St. Louis, MO) was used as a normalization control for protein loading. MDA-MB- 435s, MDA-MB-468, MDA-MB-231, MDA-MB-361, HEK-293, and NIH 3T3 cells were grown in DMEM supplemented with 10% fetal bovine serum, 100 units/mL penicillin, and 100 μg/mL streptomycin. MCFlOA cells were cultured in 1:1 mixture of Ham's F12 medium/DMEM with 2.5 mmol/L L-glutamine and supplemented with 20 ng/niL epidermal growth factor, 100 ng/niL cholera toxin, 0.01 mg/mL insulin and 500 ng/mL hydrocortisone, 5% horse serum, 100 units/mL penicillin, and 100 μg/mL streptomycin. [0064] Western blot analysis showed Survivin protein expression in all tested breast cancer cell lines with activated Stat3 (MDA-MB-231, MDA-MB-435s, and MDA-MB-468; Fig. 5A). By contrast, minimal Survivin expression was observed in breast cancer cells (MDA-MB-361) and in normal breast epithelial cells (MCF-IOA) lacking detectable Stat3 activation. Stat3 activation in each cell line was confirmed by EMSA as described above and shown in Fig. 5 A.

[0065] To investigate whether cytokine-induced Stat3 activity correlates with increased Survivin expression in breast cancer cells, cells were stimulated with IL-6 to increase Stat3 activation. For IL-6 stimulation, cells were serum-starved (DMEM supplemented with 0.1% fetal bovine serum) for 18 hours before IL-6 treatment (20 or 30 ng/mL in DMEM) for 30 minutes (EMSA) or 48 hours (Western blot analysis). Nuclear extracts and cell lysates were prepared for EMSA and Western blot analysis, respectively, as described above. [0066] IL-6 treatment of serum-starved MDA-MB-435s cells increases Stat3 DNA- binding activity within 30 minutes as detected by EMSA (Fig. 5B). In addition, both phosphor-Stat3 levels and Survivin protein expression were induced by IL-6 treatment of MDA-MB-435s cells after 48 hours (Fig. 5C). These findings indicate that constitutive and cytokine-induced Stat3 activation correlates with Survivin expression in breast cancer cell lines.

Example 6: Immunohistochemical analysis comparing tumor and nonneoplastic tissues. [0067] Clinical trials and biomarkers. Forty-five women with stage III breast carcinoma were enrolled in a 3-year clinical trial of neoadjuvant dose-dense chemotherapy with sequential doxorubicin (80 mg/m²) followed by docetaxel (100 mg/ m²) i.v. every 2 weeks for three cycles each. After neoadjuvant chemotherapy, all participants underwent surgery with either lumpectomy or mastectomy and axillary lymph node dissection. All tumors were at least 5 cm in size, 85% were ductal, 10% were lobular, and 5% had ductal and lobular features. Levels of activated tyrosine-phosphorylated Stat3 (pY-Stat3) and nine other proteins were quantified for their reported relevance to Stat activation or breast oncogenesis: pY-Src, HER2/neu, estrogen receptor (ER), progesterone receptor (PR), Ki-67, Bcl-2, BCI-X_L, epidermal growth factor receptor (EGFR), and Survivin.

[0068] Apoptosis in tissues was measured by the terminal deoxynucleotidyl transferase- mediated nick end labeling (TUNEL) assay as described above. In an attempt to identify expression patterns that correlate with response to treatment, the analysis was done in tissues obtained before and after therapy. Before treatment, tissue was obtained by incisional biopsies of tumor and nontumoral parenchyma, either from a distant ipsilateral quadrant or from the contralateral breast (majority of cases). After chemotherapy, tissue was obtained at the time of definitive surgery.

[0069] Tissue collection. To preserve the activated phosphorylation state of signal transduction proteins, tissues have to be snap frozen in liquid nitrogen within 15 minutes from the moment of interruption of blood supply to the specimen. All tissues in this study were snap frozen or fixed in 10% neutral-buffered formalin within 15 minutes to minimize antigen loss and optimize immunohistochemical detection. The presence of normal tissue or tumor was confirmed in mirror image sections of the respective samples by examination of frozen sections immediately following collection. After chemotherapy, tumor was only available from those patients with a partial pathologic response. All data obtained were entered into a web-based database for statistical correction of clinical, pathologic, and molecular data.

[0070] Heterogeneity of signal across tissue sections. An important aspect of the initial experimentation process was to determine the degree of staining variability across consecutive tissue sections. Twenty consecutive sections were prepared without discarding intervening tissue and used to perform immunohistochemistry for a signaling protein not related to the project (transforming growth factor receptor type II). Quantitative image analysis revealed minimal variation in expression intensity for the first 12 consecutive sections. Expression levels on the next set of 12 sections showed statistically significant differences when compared with the first set. Therefore, a maximum of 11 consecutive sections were used. See Diaz et al., Clin. Cancer Res. 12:20-28 (2006) which is incorporated herein by reference.

[0071] Immunohistochemistry and TUNEL procedures. Consecutive 3 μm sections were prepared without discarding intervening tissue (see above). The first section was stained with H&E and the rest of the sections were used for immunohistochemistry and TUNEL assays. For all antigens, except pY-Stat3 (see procedure for pY-Stat3 detection below), the following procedure was used. Formalin-fixed, paraffin-embedded tissue sections were dried at 37° C overnight. Sections were deparaffinized by an initial warming to 60° C, followed by two xylene changes 10 minutes each, two series of 30 dips in absolute alcohol, 30 dips in 95% alcohol, and 20 dips in deionized water. Antigen retrieval or enzyme digestion procedures were done as described by the supplier of each antibody. Slides were placed for 5 minutes in TBS/Tween and processed on a DAKO Autostainer using the Dako LSAB+ peroxidase detection kit (DAKO, Carpinteria, CA). Endogenous peroxide was blocked with 3% aqueous hydrogen peroxide followed by two 20 dips in deionized water.

[0072] The anti-EGFR monoclonal antibody clone 111.6 (Signet Pathology Systems, Dedham, MA) was applied at 1 : 100 for 30 minutes following proteinase K digestion (25 μg/mL in TBS/Tween) for 17 minutes. The rest of the antibodies were applied for 30 minutes after microwave antigen retrieval with 0.1 mol/L citrate buffer (pH 6.0; Emerson 1,100 W microwave, high to boiling, then 20 minutes on power level 5) as follows: Bcl-2 (1:40; DAKO), Bcl-x_L (1 :50; Santa Cruz Biotechnology, Santa Cruz, CA); pY-Src (1 :100; Cell Signaling Technology, Beverly, MA), Ki-67 (1 :50; Immunotech, Norcross, GA), c- ErbB-2 (1:40; HER2/neu; Signet Pathology Systems), ER and PR (1:40; BioGenex, San Ramon, CA), and Survivin (1:100; Cell Signaling Technology, Beverly, MA). The chromogen 3,3'-diaminobenzidine was used for all proteins except for Survivin, which was detected using Nova-Red (Vector Laboratories, Burlingame, CA). Survivin expression was evaluated only in samples obtained before treatment because analysis of this antigen was added at later time in this study based on microarray analyses (see above). Counterstain was done with modified Mayer's hematoxylin (Signet Laboratories, Dedham, MA). Slides were dehydrated through graded alcohol, cleared with xylene, and mounted with resinous mounting medium. Apoptosis was detected by the TUNEL assay using the Intergen Apopta G Peroxidase In situ Apoptosis detection kit (Intergen, Purchase, NY) as indicated by the supplier.

[0073] pY-Stat3 immunohistochemistry. After deparaffinizing, a two-stage pretreatment procedure was done as follows. First, antigen retrieval was done in a pressure cooker by placing a total of 600 ml deionized water in three containers, one containing the slides in citrate buffer (pH 6.0) and the other two containing only deionized water. The microwave oven (Emerson 1, 100 W) was set on high to pressurize for 12 minutes and then at power level 4 for 10 minutes. Slides were then cooled at room temperature for 30 minutes. This was followed by limited enzymatic digestion at 37° C for 5 minutes with 0.025% trypsin in 5 mmol/L Tris-Cl (pH 7.6) with 0.05% calcium chloride. At the end of the digestion, slides were rinsed with deionized water, placed in TBS/Tween for 5 minutes, drained, and framed with an ImmunoEdge pen. Hydrogen peroxide 3% was applied for 10 minutes and 3% bovine serum albumin/PBS for 10 minutes. Sections were incubated with antiphospho- Stat3 antibody (rabbit polyclonal P-State3, Cell Signaling Technologies, Beverly, MA) at 1 :400 in a humid chamber at 4° C overnight and returned to the autostainer for detection and substrate development using the Dako LS AB+ detection system and 3,3'-diaminobenzidine as chromogen. Counterstaining was done for 30 seconds with modified Mayer's hematoxylin. Sections were allowed to sit in tap water for 10 to 15 minutes and dehydrated before mounting with resinous mounting medium.

[0074] Image Analysis and quantification. The Optimas 6.5 (Media Cybernetics,

Silver Springs, MD) software was used to quantify protein expression. Regions of interest were identified on the H&E-stained slide and the same areas were marked on the consecutive sections used for each of the biomarkers and TUNEL assay. Digital images of these areas were obtained using identical magnifications (x400) and camera settings with a Leica DM microscope (Leica Microscopes, Bannockburn, IL) with neutral density 6 and 12 filters, coupled to a SPOT Digital Camera System (Diagnostic Instruments, Sterling Heights, MI) and SPOT software set as AutoGain, RGB filter color, non binning, full chip area, and adjustment factor set to 1. Before photography, Koehler epi-illumination was done. Optimal light conditions for each objective were stored as software settings to be replicated in each measurement session. Image acquisition was done after 1 hour of microscope lamp warm up. White balance was done on an area of the slide with no tissue, and background values were subtracted using a negative control slide. Images of the selected areas were stored as TIFF images. A macro was specifically set in the software to automate the process and transfer mathematical calculations to a Microsoft Excel spreadsheet. These were converted to SAS data sets for statistical analysis. Quantitative image analysis was done on all pretreatment samples and in those posttreatment samples for which tumor tissue was still identifiable (partial pathologic responders).

[0075] Immunohistochemistry and quantitative image analysis. Figure 6 shows examples of immunohistochemical staining in serial sections of nontumor (A, C, and E) and matched tumor tissues (B, D, and F) from the same patient using antibodies to HER2/neu (A and B), pY-Src (C and D), and pY-Stat3 (E and F). For quantification by digital image analysis, regions of interest were selected to include only epithelial components (nonneoplastic or carcinoma) and to eliminate mesenchymal areas. Each of the markers was evaluated in its corresponding cellular compartment (nucleus, cytoplasm, or membrane). In the case of predominantly nuclear markers like pY-Stat3 (A, B, and C), thresholds were set to discriminate between the brown color of 3,3'-diaminobenzidine (or the red color of Nova- Red for Survivin) and the blue color of hematoxylin in negative nuclei. For membranous markers like HER2/neu (D, E, and F), the 3,3'-diaminobenzidine signal was measured as a percentage of the total cellular area after subtraction from the region of interest of both nuclear and cytoplasmic components. For predominantly cytoplasmic markers like pY-Src (G, H, and I), the brown 3,3'-diaminobenzidine signal was quantified in the total cellular area of the selected region of interest after subtraction of the blue nuclear area. In all three situations, positive signals were reported as an index reflecting the optical intensity of the marker in relationship to the total optical intensity of the region of interest (average of three measurements) . [0076] HER2/neu clinical testing. HER/2neu status was also assessed in all tumors at an independent laboratory using a Food and Drug Administration-approved immunohistochemistry procedure. Results of this test are reported negative when intensity scores are 0 or 1+ and positive when the intensity score is 3+. Cases with intermediate intensity score of 2+ are further evaluated by fluorescence in situ hybridization. Using this method, 11 cases (24.4%) were positive, 31 (68.9%) were negative, and 3 (6.7%) remained undetermined. These percentages are consistent with previous reports. In the group of tumors that were positive by the Food and Drug Administrative-approved test, the average HER2/neu index calculated by computerized image analysis in this study was 50 and the SD 14.5. In the group of tumors that were negative by the Food and Drug Administrative- approved test, the average HER2/neu index by computerized image analysis was 35.5 and the SD was 30. The difference was not statistically significant but suggested a similar trend of HER2/neu detection with both methods.

[0077] Statistical considerations. Before statistical evaluation, pathologic response was classified as either complete pathologic response (CPR) or partial pathologic response (PPR) based on the size of residual tumor after treatment (complete pathologic response if 0 cm, partial pathologic responses if >0 cm). Biomarker values were analyzed by descriptive statistics, including mean, SD, and range (minimal-maximal), values in all patients as a group and in the complete pathologic response and partial pathologic response groups separately both in nonneoplastic tissue and in tumor tissue. Comparison of pretreatment samples between the complete pathologic response and partial pathologic response groups was examined using the Student's t test if the data followed a normal distribution and the Wilcoxon Mann- Whitney test if the normality assumption was not met. In addition, pretreatment to posttreatment changes were evaluated using the paired t test in the partial pathologic response group. Correlations between pretreatment values were assessed by the Spearman's rank correlation coefficient. No adjustments were made for multiple testing owing to the exploratory nature of this study. All tests were two-sided and declared significant at the 5% level.

[0078] Table 2 summarizes the descriptive statistics of the immunohistochemical quantitative analysis in nonneoplastic and tumor tissues by pathologic response for pretreatment and posttreatment. Of the 45 patients included in the clinical trial, 12 (27%) showed a complete pathologic response and 33 (73%) showed a partial pathologic response. Statistical analysis of the immunohistochemical data was done in all 45 patients as a group and in the complete pathologic response and partial pathologic response groups separate, as described below.

[0079] Molecular biomarker levels in tumor versus nonneoplastic tissues. Table 3 shows that in tissues obtained before chemotherapy treatment, levels of pY-Stat3, pY-Src, and Ki-67 were significantly higher in tumors than in nonneoplastic tissues, both when all tumors were analyzed as a group without regard to pathologic response (P ≤O.OOl) and when the analysis was done in the complete pathologic response and partial pathologic response groups separately. In contrast, pretreatment levels of ER in the complete pathologic response group, and Bcl-2 in both complete pathologic response and partial pathologic response groups, were significantly higher in nonneoplastic tissues than in tumors (compare Tables 2 and 3). Similar relationships were found in tissues after chemotherapy treatment, with higher levels of pY-Stat3, pY-Src, and Ki-67 in tumors and higher levels of Bcl-2 in nonneoplastic tissues.

[0080] Correlation between biomarker levels in tumors. No significant correlations were found among biomarkers in nonneoplastic tissues, either in the pretreatment or in the posttreatment groups (data now shown). In tumors, however, the following significant statistical correlations among molecular biomarkers were observed (Table 4).

[0081] Pretreatment. When pretreatment tumors were analyzed as a group, without regard to pathologic response, ER correlated with EGFR; PR with Bcl-2 and Bcl-xi; pY-Src with Bcl-2, Bcl-xi, EGFR, and Ki-67; and pY-Stat3 with HER2/neu and Survivin. In the complete pathologic response group, Bcl-2 correlated with BcI-X₁ and PR, and pY-Stat3 with HER2/neu and pY-Src. In the partial pathologic response group, EGFR correlated with ER and TUNEL; pY-Src with Bcl-2, BcI-X₁ and PR, and EGFR; and pY-Stat3 with ER and Survivin. When pretreatment tumor values in the complete pathologic response group were compared with those of the partial pathologic response, only pY-Stat3 was found to have a statistically significant correlation with response to therapy (P = 0.028). Levels of pY-Stat3 were lower in tumors of patients who showed complete pathologic response, suggesting that higher levels of activated Stat3 made tumors less responsive to the treatment. Taking all of the data into consideration, the most statistically significant correlation with clinical relevance among the pretreatment biomarkers was between pY-Stat3 and Survivin levels in the partial pathologic response group (P = 0.001).

[0082] Posttreatment. In the posttreatment partial pathologic response group, a direct correlation between HER2/neu and Ki-67 and an inverse correlation between EGFR and pY-

Src were observed.

[0083] Pretreatment versus posttreatment (partial pathologic responders only). Levels of Ki-67 were higher in pretreatment samples than in posttreatment samples (P < 0.02), and levels of Bcl-2 were higher in posttreatment samples than in pretreatment samples (P < 0.02).

Example 7: Stat3 activation and Survivin expression correlate with response to chemotherapy.

[0084] The association between constitutive Stat3 activation and Survivin expression in primary breast tumors was investigated. Tissue specimens were analyzed by immunohistochemical staining of formalin-fixed, paraffin-embedded sections using phosphor-Stat3 or Survivin antibodies. Moderate to strong predominantly nuclear staining was observed for phosphor-Stat3 and Survivin in a majority of the tumor specimens but not in normal breast epithelial cells (Fig. 7). Importantly, a statistically significant positive correlation (P = 0.001) was observed between elevated phosphorylated Stat3 levels and

Survivin expression in 33 of the 45 breast cancer patients who displayed a partial pathologic response to this neoadjuvant chemotherapy regimen. Thus, high levels of phosphorylated

Stat3 and Survivin expression correlate with invasive breast cancer and resistance to chemotherapy.

[0085] Correlation of molecular biomarkers with response to chemotherapy. As described above, Stat3 activity was measured by both immunohistochemistry and EMSA in the pretreatment tumor and matched normal tissue samples. The other markers, including phosphorylated-Src, HER2/neu, EGF-R, ER, Ki-67, apoptotic index (TUNEL), Bcl-2, Bcl-x_L, and Survivin, were measured by immunohistochemistry. Of all the molecular biomarkers examined, only activated phosphorylated Stat3 measured by immunohistochemistry was statistically associated (p=0.028) with decreased likelihood of pathologic complete response to the chemotherapy. Our statistical analysis estimates that the odds of a pathological complete response decreases by a factor of 0.93 for every unit increase in phosphorylated Stat3 levels (Table 5). Table 5. Molecular Biomarkers that Predict Pathologic Complete Response

Molecular Estimated Standard Odds

Biomarker Coefficient Error P-Value Ratio (95% Cl) p-Stat3 -0.0704 0.0320 0.0277 0.93 (0.88-0.99) p-Src 0.0410 0.0241 0.0885 1.04 (0.99-1.09)

Survivin 0.0183 0.0233 0.4301 1.02 (0.97-1.07)

Her2/neu -0.0206 0.0206 0.3177 0.98 (0.94-1.02)

EGF-R -0.0183 0.0210 0.3841 0.98 (0.94-1.02)

ER 0.0296 0.0290 0.3073 1.03 (0.97-1.09)

Ki-67 0.0263 0.0235 0.2619 1.03 (0.98-1.08)

Tunel -0.0192 0.0217 0.3766 0.98 (0.94-1.02)

Bcl-2 -0.0471 0.0331 0.1550 0.95 (0.89-1.02)

Bcl-xL 0.00700 0.0167 0.6749 1.01 (0.98-1.04)

[0086] Figure 8 illustrates the distribution of phosphorylated Stat3 levels measured by immunohistochemistry in the complete versus partial pathologic response groups. Although statistical comparison of the medians for each of the distributions was marginally significant (p=0.057 by the Wilcoxon rank sum test), examination of the data indicates that the levels of phosphorylated Stat3 tend to be greater in the partial pathological response group. Example 8: Stat3 activation in non-small cell lung cancer.

[0087] Tissue arrays and immunohistochemistry. The tissue arrays from stage 1 NSCLC have been described by Zheng et al., Chest 128:308-16 (2005), which is incorporated herein by reference. The samples collected for our tissue microarray construction met the following requirements: (a) diagnosis of stage 1 NSCLC without preoperative radiation or chemotherapy, (b) surgically resected specimens formalin fixed and embedded in paraffin block, and (c) adequate tumor tissue in size for at least three tissue cores. Immunostaining for phosphorylated Stat3 was done as described above using a rabbit anti-human polyclonal antibody (phosphotyrosine-Stat3 705; Cell Signaling Technology, Beverly, MA); as negative controls, rabbit immunoglobulins (Vector, Burlingame, CA) were used as a primary antibody. Immunostaining for pEGFR was done using a rabbit anti-human polyclonal antibody (phosphotyrosine-EGFR 845, Cell Signaling Technology, Beverly, MA). [0088] Immunohistochemical staining of the slides was reviewed and scored. For phosphorylated Stat3, the score system included counting the percentage of nuclear phosphorylated Stat3-stained tumor cells and estimating the intensity of nuclear phosphorylated Stat3-stained tumor cells in a semiquantitative manner. Because transcriptional activity of Stat3 requires nuclear localization to regulate gene expression, we chose nuclear staining instead of cytoplasmic staining. For pEGFR, the score system included counting the percentage of positively stained tumor cells and estimating the intensity staining in a semiquantitative manner. Intensity was classified as 0 to 3 (0 = no staining, 1 = weak staining, 2 = medium, and 3 = strong staining). For triplicate samples, mean values of the percentage of staining, the intensity, and a composite score (percentage multiplied by intensity, range 0-300) were derived.

[0089] Patterns of activated, phosphorylated Stat3 were characterized using high-density tissue microarrays using surgical resected NSCLC. Primary NSCLC from 176 patients were arrayed in triplicate. A phosphotyrosine 705-Stat3 antibody was used to investigate the levels and cytologic locations of activated Stat3 (termed pStat3) in tumor specimens as described by Mora et al., Cancer Res 62:6659-66 (2002), which is incorporated herein by reference. Tyrosine 705 on Stat3 is phosphorylated by upstream tyrosine kinases, such as EGFR, and this allows for dimer formation, translocation to the nucleus, and DNA binding. Because transcriptional activity of Stat3 requires nuclear localization to regulate gene expression, nuclear staining was evaluated. For each tumor, the mean composite nuclear phosphorylated Stat3 score (intensity multiplied by percentage of tumor cells stained in the three cores) was recorded and correlated with clinical and pathologic variables. A description of the patient characteristics is shown in Table 6.

Results of phosphorylated Stat3 staining and correlations with patient characteristics are shown in Table 7.

[0090] These results show nuclear pStat3 expression in 54% of tumors (composite score range 0-222). Higher pStat3 expression was found in patients with limited smoking history (P = 0.02) and in smaller tumors (P < 0.0001). A trend toward higher pStat3 expression was identified in tumors with bronchioloalveolar carcinoma or adenocarcinoma histology compared with either squamous cell or large cell histology (P = 0.09). No relationship was identified between pStat3 and either overall or disease-free survival following surgical resection.

[0091] These results are similar to those reported earlier for pEGFR. As reported by Zheng et al, Chest 128:308-16 (2005), expression of pEGFR was noted in 51% of tumor tissues with higher expression in patients with low smoking history and in smaller tumors. Consistent with laboratory studies showing EGFR signaling can activate Stat3, a very strong correlation was identified between pEGFR expression and pStat3 expression in this patient cohort (Spearman's/? = 0.55, P< 0.0001). Example 9: Stat3 activation promotes tumor survival.

[0092] Phosphorylated Stat3 expression was next correlated with apoptosis within the primary tumor. Tumor apoptosis was evaluated by detecting apoptotic cells and apoptotic bodies using in situ labeling with an ApopTag Plus Peroxidase In situ Apoptosis Detection kit (Chemicon International, Temecula, CA). This technique detects DNA fragmentation due to

apoptosis at the single cell level. The 3' -OH termini generated by DNA end nicking were

labeled with modified nucleotides by terminal deoxynucleotidyl transferase, which more selectively detects apoptotic cells rather than necrotic cells. The incorporation of these modified nucleotides was assayed by a specific antibody in immunohistochemical analysis. The scoring system for ApopTag involved counting the percentage of positively stained tumor cells, estimating the intensity staining of a semiquantitative manner as before, generating a composite score and taking the mean value from the three cores. Consistent with the role of Stat3 in promoting tumor cell survival, a negative correlation between phosphorylated Stat3 and tumor apoptosis was found (Spearman's/? = -0.19, P = 0.01). This result is consistent with less apoptosis in tumors having higher amounts of phosphorylated Stat3.

Example 10: Stat3 activation correlates with sensitivity to EGFR inhibitors. [0093] Cell lines and cell culture. Human NSCLC cell lines were purchased from American Type Culture Collection (ATCC; Manassas, VA). H3244 cells were provided by Dr. Pasi Janne and grown as described in Fujishita et al, Oncology 64:399-406 (2003), which is incorporated herein by reference. All other cells were grown in RPMI 1640 supplemented with 2 mmol/L L-glutamine (Santa Cruz Biotechnology, Santa Cruz, CA) and 5% bovine calf serum (Hyclone). Subconfluent cells were prepared for protein extracts as detailed below. Cell numbers were determined by counting with a hemocytometer. Gefitinib (ZD 1839) was provided by Astra Zeneca (Wilmington, DE).

[0094] Cytotoxicity assays. Cytotoxicity assays [3-(4,5-dimethylthiazol-2-yl)-2,5- diphenyltetrazolium bromide] were done according to the recommendations of the manufacturer (Roche, Indianapolis, IN). Cells were counted and 5 x 10⁴ cells were placed into single wells in a 96-well plate. Cells were grown as above and exposed to the indicated agents as described. Cell viability was assessed following 72 hours. Data presented represents two separate experiments with eight data points per condition. Data were expressed as mean of eight data points.

[0095] Protein expression analysis. Cell lysates were normalized for total protein content (50 μg) and subjected to SDS-PAGE as described by Song et al., Oncogene 22:4150-65 (2003). Primary antibodies used in these studies consisted of Stat3 (Transduction Laboratories; BD Biosciences, Mississauga, Ontario, Canada), pStat3 Y705 (Cell Signaling Technology), cleaved poly (ADP-ribose) polymerase (Cell Signaling Technology, Beverly, MA), McI-I (Santa Cruz Biotechnology, Santa Cruz, CA), Bcl-2 (DAKO, Carpinteria, CA),

Bcl-xL (PharMingen, San Diego, CA), and β-actin (Sigma- Aldrich, Milwaukee, WI).

Detection of proteins was accomplished using horseradish peroxidase-conjugated secondary antibodies and enhanced chemiluminescence purchased through Amersham Biosciences (Pittsburgh, PA).

[0096] The level of activated Stat3 was assessed in cell lines harboring mutant EGFR and cell lines with wild-type EGFR. See Haura et al., Clin. Cancer Res. 11 :8288-8294 (2005), which is incorporated herein by reference. H3255, H1650, and H1975 (cell lines that contain mutations of EGFR), along with H460, H358, H1299, and A549 (NSCLC cell lines with wild-type EGFR), were exposed to increasing concentrations of gefitinib and cell viability was assayed. As shown in Fig. 9A, mutant EGFR cells were sensitive to gefitinib with an approximate IC₅₀ of 100 nniol/L, whereas wild-type cell lines were resistant to gefitinib (IC₅₀ > 10 μmol/L). To confirm that gefitinib sensitivity correlated with increased Stat3 activation, whole cell proteins from untreated cells were evaluated for phosphorylated Stat3 and total Stat3. The mutation status and sensitivity of cell lines to gefitinib correlated with the level of whole cell phosphorylated Stat3 activity because H3255, H1650, and H1975 cells have markedly higher levels of phosphorylated Stat3 compared with the other cell lines (Fig. 9B). [0097] To evaluate whether inhibition of EGFR tyrosine kinase activity by gefitinib affects downstream Stat3 activity, NSCLC cells with distinct EGFR mutations were exposed to increasing doses of gefitinib for 24 hours and total proteins were evaluated for phosphorylated Stat3 and total Stat3 (Fig. 9C). In H3255 cells with the L858R mutation in EGFR, inhibition of EGFR tyrosine kinase by gefitinib resulted in a complete inhibition of phosphorylated Stat3. This suggested that loss of Stat3 activity contributes to gefitinib- mediated apoptosis. However, in Hl 650 cells with the L746-p753 deletion mutation in EGFR or in gefitinib-sensitive Hl 975 cells, gefitinib had no effect on phosphorylated Stat3. Gefitinib inhibition of EGFR function was confirmed by analysis of pTyr 1068 EGFR levels. Phosphorylated EGFR was down-regulated in all the cells tested. Therefore, despite enhanced levels of phosphorylated Stat3 in cells harboring EGFR mutations, the ability of gefitinib to down-regulate Stat3 activity seemed to be cell line dependent. [0098] To determine if Stat3 plays a role in survival in EGFR mutant cells, antisense Stat3 oligonucleotides were used to down-regulate Stat3 as described above. Hl 650 and Hl 975 cells exposed to antisense Stat3 show decreased Stat3 protein levels as well as cleavage of poly(ADP-ribose) polymerase indicative of apoptosis (Fig. 9D). No apoptosis was observed in Hl 299 cells despite Stat3 knock-down. Levels of the Bcl-2 family proteins Bcl-xL, McI-I, and Bcl-2 were reduced in cells treated with antisense Stat3 consistent with the known role of Stat3 in regulating these proteins (Fig. 9E). These results suggest that Stat3 communicates a survival signal from mutant EGFR to downstream survival proteins. [0099] All publications and references, including but not limited to patents and patent applications, cited in this specification are herein incorporated by reference in their entirety as if each individual publication or reference were specifically and individually indicated to be incorporated by reference herein as being fully set forth. Any patent application to which this application claims priority is also incorporated by reference herein in its entirety in the manner described above for publications and references.

[00100] It will be obvious to those of ordinary skill in the art that variations in the preferred devices and methods may be used and that it is intended that the invention may be practiced otherwise than as specifically described herein. Accordingly, this invention includes all modifications encompassed within the spirit and scope of the invention as defined by the claims that follow.

Claims

We claim:

1. A method of predicting responsiveness of cancer cells to a chemotherapeutic agent, comprising: a) measuring the level of phosphorylated Stat in a cancer cell; and b) comparing the level of phosphorylated Stat in the cancer cell to the level of phosphorylated Stat in a control, wherein the level of phosphorylated Stat in the cancer cell as compared to the control is predictive of responsiveness, to a chemotherapeutic agent.

2. The method of claim 1, wherein the cancer cell is obtained from a subject.

3. The method of claim 2, wherein the subject is human.

4. The method of claim 1, wherein the cancer cell is selected from the group consisting of a breast cancer cell, a lung cancer cell, an ovarian cancer cell, a head and neck cancer cell, a melanoma cell, a prostate cancer cell, a multiple myeloma cell, a lymphoma cell, a leukemia cell, a gastric cancer cell, an ovary cancer cell, a colon cancer cell, and a pancreatic cancer cell.

5. The method of claim 4, wherein the cancer cell is a breast cancer cell.

6. The method of claim 4, wherein the cancer cell is a non-small cell lung cancer cell.

7. The method of claim 1, wherein the chemotherapeutic agent is selected from the group consisting of alkylating agents, nitrosoureas, antimetabolites, anthracyclines, topoisomerase inhibitors, taxanes, vinca alkaloids, and tyrosine kinase inhibitors.

8. The method of claim 1, wherein the chemotherapeutic agent is a taxane.

9. The method of claim 8, wherein the taxane is docetaxel.

10. The method of claim 1, wherein the chemotherapeutic agent is a tyrosine kinase inhibitor.

11. The method of claim 10, wherein the tyrosine kinase inhibitor is selected from the group consisting of Stat inhibitors and EGFR inhibitors.

12. The method of claim 11 , wherein the EGFR tyrosine kinase inhibitor is selected from the group consisting of gefitinib and erlotinib.

13. The method of claim 1, wherein the chemotherapeutic agent is doxorubicin.

14. The method of claim 1, wherein the level of phosphorylated Stat is increased in the cancer cell relative to the control and is predictive of responsiveness to tyrosine kinase inhibitors.

15. The method of claim 1, wherein the level of phosphorylated Stat is not increased in the cancer cell relative to the control and is predictive of responsiveness to chemotherapeutic agents that induce apoptosis.

16. The method of claim 1, wherein the cancer cell is obtained by a method selected from the group consisting of a biopsy and surgical resection.

17. The method of claim 1, wherein the level of phosphorylated Stat is measured by a method selected from the group consisting of immunohistochemistry, electrophoretic mobility shift assay, Western blot and ELISA.

18. The method of claim 1, wherein the predicted responsiveness is a complete pathologic response.

19. The method of claim 1, wherein the predicted responsiveness is a complete clinical response.

20. The method of claim 1, wherein the control is a non-cancerous cell.

21. The method of claim 20, wherein the non-cancerous cell is obtained from the same subject as the cancer cell.

22. The method of claim 1, wherein the control is a cancer cell.

23. The method of claim 1, wherein the control has a known responsiveness to chemotherapeutic agents.

24. The method of claim 1, wherein the Stat is selected from the group consisting of Statl, Stat3, Stat5a and Stat5b.

25. The method of claim 1, wherein the Stat is Stat3.

26. A method of predicting responsiveness of a cancer cell to a chemotherapeutic agent, comprising: a) measuring the level of expression of Survivin in a cancer cell; and b) comparing the level of expression of Survivin in the cancer cell from step (a) to the level of expression of Survivin in a control, wherein the level of expression of Survivin in the cancer cell as compared to the control is predictive of responsiveness to a chemotherapeutic agent.

27. The method of claim 26, wherein the level of expression of Survivin is measured by a method selected from the group consisting of immunohistochemistry, Western blot, ELISA, microarray, rt-PCR, and Northern blot.

28. A method for selecting a chemotherapeutic agent for a subject diagnosed with cancer comprising: a) measuring the level of phosphorylated Stat in a cancer cell from the subject; b) comparing the level of phosphorylated Stat in the cancer cell to the level of phosphorylated Stat in a control; and c) selecting a chemotherapeutic agent for the subject based on the level phosphorylated Stat in the cancer cell as compared to the level in the control.

29. The method of claim 28, wherein an increased level of phosphorylated Stat in the cancer cell as compared to a control cell is indicative of sensitivity of the cancer cell to EGFR tyrosine kinase inhibitors.

30. The method of claim 28, wherein an increased level of phosphorylated Stat in the cancer cell as compared to a control cell is indicative of sensitivity of the cancer cell to Src inhibitors.

31. The method of claim 28, wherein an increased level of phosphorylated Stat in the cancer cell as compared to a control cell is indicative of sensitivity of the cancer cell to Stat inhibitors.

32. The method of claim 28, wherein an increased level of phosphorylated Stat in the cancer cell as compared to a control cell is indicative of reduced sensitivity of the cancer cell to taxanes.

33. The method of claim 28, wherein an increased level of phosphorylated Stat in the cancer cell as compared to a control cell is indicative of reduced sensitivity of the cancer cell to anthracyclines.

34. A method of treating a subject diagnosed with cancer comprising: a) measuring the level of phosphorylated Stat in a cancer cell from the subject; b) comparing the level of phosphorylated Stat in the cancer cell to the level of phosphorylated Stat in a control; c) selecting a chemotherapeutic agent for the subject based on the level phosphorylated Stat in the cancer cell as compared to the level in the control; and d) administering an effective amount of the chemotherapeutic agent to the subject to treat the cancer.

35. A method of modulating Survivin-dependent apoptosis in a cancer cell comprising contacting the cell with an effective amount of a Stat inhibitor.

36. A method of treating a subject diagnosed with cancer comprising: a) measuring the level of expression of Survivin in a cancer cell from the subject; b) comparing the level of expression of Survivin in the cancer cell to the level of expression of Survivin in a control; c) selecting a chemotherapeutic agent for the subject based on the level of expression of Survivin in the cancer cell as compared to the level of expression of Survivin in the control; and d) administering an effective amount of the chemotherapeutic agent to the subject to treat the cancer.

37. A method for selecting a chemotherapeutic treatment for a subject diagnosed with cancer comprising: a) measuring the level of expression of Survivin in a cancer cell from the subject; b) comparing the level of expression of Survivin in the cancer cell to the level of expression of Survivin in a control; and c) selecting a chemotherapeutic agent for the subject based on the level of expression of Survivin in the cancer cell as compared to the level of expression of Survivin in the control.

38. A kit for predicting responsiveness of cancer cells to a chemotherapeutic agent comprising an antibody capable of binding phosphorylated Stat.

39. A kit for predicting responsiveness of cancer cells to a chemotherapeutic agent comprising an antibody capable of binding Survivin.

40. A kit for performing the method of claim 26 comprising at least two oligonucleotides capable of amplifying a polynucleotide encoding Survivin.

41. A kit for performing the method of claim 26 comprising an oligonucleotide capable of hybridizing to Survivin mRNA.