(12) INTERNATIONAL APPLICATION PUBLISHED UNDER THE PATENT COOPERATION TREATY (PCT)
`
`(19) World Intellectual Property Organization ”Va | I
`International Bureau
`
`I
`
`
`
`(43) International Publication Date
`10 April 2008 (10.04.2008)
`
`International Patent Classification:
`A61K 39/00 (2006.01)
`
`International Application Number:
`PCT/US2007/020889
`
`International Filing Date:
`28 September 2007 (28.09.2007)
`
`(25)
`
`Filing Language:
`
`Publication Language:
`
`English
`
`English
`
`Priority Data:
`60/847,904
`60/886,260
`60/942,542
`
`29 September 2006 (29092006)
`23 January 2007 (23.01.2007)
`7 June 2007 (07.06.2007)
`
`US
`US
`US
`
`(84)
`
`Applicant (for all designated States except US): ON-
`COMEI) PHARMACEUTICALS, INC. [US/US]; 800
`Chesapeake Drive, Redwood City, CA 94063 (US).
`
`Inventors; and
`Inventors/Applicants (for US only): GURNEY, Austin
`[US/US]; 496 Diamond Street, San Francisco, CA 94114
`(US). HOEY, Timothy [US/US]; 200 Darrell Road, Hills,
`borough, CA 94010 (US). SATYAL,Sanjeev [IN/US]; 965
`Buckland Avenue, San Carlos, CA 94070 (US). AXEL-
`ROD, Fumiko [US/US]; 3780 Wright Place, Palo Alto, CA
`94306 (US).
`
`(74)
`
`Agents: MCCABE, Kevin, W. et al.; Sterne, Kessler,
`Goldstein & Fox P.L.L.C., 1100 New York Avenue, N.W.,
`Washington, DC 20005—3934 (US).
`
`(10) International Publication Number
`
`WO 2008/042236 A2
`
`(81)
`
`Designated States (unless otherwise indicated, for every
`kind of national protection available): AE, AG, AL, AM,
`AT, AU, AZ, BA, BB, BG, BH, BR, BW, BY, BZ, CA, CH,
`CN, CO, CR, CU, CZ, DE, DK, DM, DO, DZ, EC, EE, EG,
`ES, FI, GB, GD, GE, GH, GM, GT, HN, HR, HU, ID, IL,
`IN, IS, JP, KE, KG, KM, KN, KP, KR, KZ, LA, LC, LK,
`LR, LS, LT, LU, LY, MA, MD, ME, MG, MK, MN, MW,
`MX, MY, MZ, NA, NG, NI, NO, NZ, OM, PG, PH, PL,
`PT, RO, RS, RU, SC, SD, SE, SG, SK, SL, SM, SV, SY,
`TJ, TM, TN, TR, TT, TZ, UA, UG, US, UZ, VC, VN, ZA,
`ZM, ZW.
`
`Designated States (unless otherwise indicated, for every
`kind of regional protection available): ARIPO (BW, GH,
`GM, KE, LS, MW, MZ, NA, SD, SL, SZ, TZ, UG, ZM,
`ZW), Eurasian (AM, AZ, BY, KG, KZ, MD, RU, TJ, TM),
`European (AT, BE, BG, CH, CY, CZ, DE, DK, EE, ES, Fl,
`FR, GB, GR, HU, IE, IS, IT, LT, LU, LV, MC, MT, NL, PL,
`PT, RO, SE, SI, SK, TR), OAPI (BF, BJ, CF, CG, CI, CM,
`GA, GN, GQ, GW, ML, MR, NE, SN, TD, TG).
`
`Published:
`
`without international search report and to be republished
`upon receipt of that report
`with (an) indication(s) in relation to deposited biological
`material furnished under Rule 13bis separately from the
`description
`with sequence listing part of description published sepa—
`rately in electronicform and available upon requestfrom
`the International Bureau
`
`(51)
`
`(21)
`
`(22)
`
`(26)
`
`(30)
`
`(71)
`
`(72)
`(75)
`
`
`
`W02008/042236A2|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
`
`(54) Title: COMPOSITIONS AND METHODS FOR DIAGNOSING AND TREATING CANCER
`
`(57) Abstract: An isolated antibody that specifically binds to an extracellular domain of human DLL4 and affects growth of a tumor
`comprising cancer stem cells is described. Also described is a method of treating cancer comprising administering a therapeutically
`effective amount of an antirDLL4 antibody.
`
`

`

`WO 2008/042236
`
`PCT/U52007/020889
`
`Compositions and Methods for Diagnosing and Treating Cancer
`
`Field
`
`DESCRIPTION OF THE INVENTION
`
`[001]
`
`The present invention relates to the field of oncology and provides novel
`
`compositions and methods for diagnosing and treating cancer. The present invention
`
`provides antibodies against a cancer stem cell marker for the diagnosis and treatment of solid
`
`tumors .
`
`Background
`
`[002]
`
`Cancer is one of the leading causes of death in the developed world, with
`
`over one million people diagnosed with cancer and 500,000 deaths per year in the United
`
`States alone. Overall it is estimated that more than 1 in 3 people will develop some form of
`
`cancer during their lifetime. There are more than 200 different types of cancer, four of
`
`which—breast, lung, colorectal, and prostate—account for over half of all new cases (Jemal
`
`et al., 2003, CancerJ. Clin. 5325—26).
`
`[003]
`
`Breast cancer is the most common cancer in women, with an estimate 12%
`
`of women at risk of developing the disease during their lifetime. Although mortality rates
`
`have decreased due to earlier detection and improved treatments, breast cancer remains a
`
`leading cause of death in middle-aged women, and metastatic breast cancer is still an
`
`incurable disease. On presentation, most patients with metastatic breast cancer have only one
`
`or two organ systems affected, but as the disease progresses, multiple sites usually become
`
`involved. The most common sites of metastatic involvement are locoregional recurrences in
`
`the skin and soft tissues of the chest wall, as well as in axilla and supraclavicular areas. The
`
`most common site for distant metastasis is the bone (30 - 40% of distant metastasis), followed
`
`by the lungs and liver. And although only approximately 1—5% of women with newly
`
`1
`
`

`

`WO 2008/042236
`
`PCT/US2007/020889
`
`diagnosed breast cancer have distant metastasis at the time of diagnosis, approximately 50%
`
`of patients with local disease eventually relapse with metastasis within five years. At present
`
`the median survival from the manifestation of distant metastases is about three years.
`
`[004]
`
`Current methods of diagnosing and staging breast cancer include the tumor-
`
`node—metastasis (TNM) system that relies on tumor size, tumor presence in lymph nodes, and
`
`the presence of distant metastases (American Joint Committee on Cancer: AJCC Cancer
`
`Staging Manual. Philadelphia, Pa.: Lippincott-Raven Publishers, 5th ed., 1997, pp 171-180;
`
`Harris, J R: “Staging of breast carcinoma” in Harris, J. R., Hellman, 8.; Henderson, I. C.,
`Kinne D. W. (eds): Breast Diseases. Philadelphia, Lippincott, 1991). These parameters are
`
`used to provide a progndsis and select an appropriate therapy. The morphologic appearance
`ofthe tumor can also be assessed but because tumors with similar histopathologic appearance
`
`can exhibit significant clinical variability, this approach has serious limitations. Finally
`
`assays for cell surface markers can be used to divide certain tumors types into subclasses.
`
`For example, one factor considered in the prognosis and treatment of breast cancer is the
`
`presence of the estrogen receptor (ER) as EEK-positive breast cancers typically respond more
`
`readily to hormonal therapies such as tamoxifen or aromatase inhibitors than ER—negative
`
`tumors. Yet these analyses, though useful, are only partially predictive of the clinical
`
`behavior of breast tumors, and there is much phenotypic diversity present in breast cancers
`
`that current diagnostic tools fail to detect and current therapies fail to treat.
`
`[005]
`
`Prostate cancer is the most common cancer in men in the developed world,
`
`representing an estimated 33% of all new cancer cases in the U.S., and is the second most
`
`frequent cause of death (Jemal et a1., 2003, CA Cancer J. Clin. 53:5-26). Since the
`introduction of the prostate specific antigen (PSA) blood test, early detection of;prostate
`
`cancer has dramatically improved survival rates; the five year survival rate for patients with
`
`local and regional stage prostate cancers at the time of diagnosis is nearing 100%. Yet more
`
`

`

`WO 2008/042236
`
`PCT/US2007/020889
`
`than 50% of patients will eventually deveIOp locally advanced or metastatic disease
`
`(Muthuramalingam et al., 2004, Clin. Oncol. 162505-16).
`
`[006]
`
`Currently radical prostatectomy and radiation therapy provide curative
`
`treatment for the majority of localized prostate tumors. However, therapeutic options are
`
`very limited for advanced cases. For metastatic disease, androgen ablation with luteinising
`
`hormone-releasing hormone (LHRH) agonist alone or in combination with anti—androgens is
`
`the standard treatment. Yet despite maximal androgen blockage, the disease nearly always
`
`progresses with the majority developing androgen-independent disease. At present there is
`
`no uniformly accepted treatment for hormone refractory prostate cancer, and
`
`chemotherapeutic regimes are commonly used (Muthuramalingam et al., 2004, Clin. Oncol.
`
`16:505-16; Trojan et al., 2005, Anticancer Res. 25:551-61).
`
`[007]
`
`Colorectal cancer is the third most common cancer and the fourth most
`
`frequent cause of cancer deaths worldwide (Weitz et a1., 2005, Lancet 365:153—65).
`
`Approximately 5-10% of all colorectal cancers are hereditary with one of the main forms
`
`being familial adenomatous polyposis (FAP), an autosomal dominant disease in which about
`
`80% of affected individuals contain a germline mutation in the adenomatous polyposis coli
`
`(APC) gene. Colorectal carcinomas invade locally by circumferential growth and elsewhere
`
`by lymphatic, hematogenous, transpen'toneal, and perineural spread. The most common site
`
`of extra]yrnphatic involvement is the liver, with the lungs the most frequently affected extra-
`
`abdominal organ. Other sites of hematogenous spread include the bones, kidneys, adrenal
`
`glands, and brain.
`
`[008]
`
`The current staging system for colorectal cancer is based on the degree of
`
`tumor penetration through the bowel wall and the presence or absence of nodal involvement.
`Thisstaging system is defined by three major Duke’s classifications: Duke’s A disease is
`
`confined to submucosa layers of colon or rectum; Duke’s B disease has tumors that invade
`
`

`

`WO 2008/042236
`
`PCT/US2007/020889
`
`through the muscularis propria and may penetrate the wall of the colon or rectum; and Duke’s
`
`C disease includes any degree of bowel wall invasion with regional lymph node metastasis.
`
`While surgical resection is highly effective for early stage colorectal cancers, providing cure
`
`rates of 95% in Duke’s A patients, the rate is reduced to 75% in Duke’s B patients and the
`
`presence of positive lymph node in Duke’s C disease predicts a 60% likelihood of recurrence
`
`within five years. Treatment of Duke’s C patients with a post surgical course of
`
`chemotherapy reduces the recurrence rate to 40%-5 0% and is now the standard of care for
`
`these patients.
`
`’
`
`‘
`
`[009]
`
`Lung cancer is the most common cancer worldwide, the third most
`
`commonly diagnosed cancer in the United States, and by far the most frequent cause of
`
`cancer deaths (Spiro et al., 2002, Am. J. Respir. Crit. Care Med. 166:1]66—96; Jemal et al.,
`
`2003, CA CancerJ. Clin. 53:5-26). Cigarette smoking is believed responsible for an ’
`
`estimated 87% of all lung cancers making it the most deadly preventable disease. Lung
`
`cancer is divided into two major types that account for over 90% of all lung cancers: small
`
`cell lung cancer (SCLC) and non—small cell lung cancer (NSCLC). SCLC accounts for 15—
`
`20% of cases and is characterized by its origin in large central airways and histological
`
`composition of sheets of small cells with little cytoplasm. SCLC is more aggressive than
`
`NSCLC, growing rapidly and metastasizing early. NSCLC accounts for 80—85% of all cases
`
`and is further divided into three major subtypes based on histology: adenocarcinoma,
`
`squamous cell carcinoma (epidermoid carcinoma), and large cell undifferentiated carcinoma.
`
`[010]
`
`Lung cancer typically presents late in its course, and thus has a median
`
`survival of only 6-12 months after diagnosis and an overall 5 year survival rate of only 5-
`
`10%. Although surgery offers the best chance of a cure, only a small fraction of lung cancer
`
`patients are eligible with the majority relying on chemotherapy and radiotherapy. Despite
`
`attempts to manipulate the timing and dose intensity of these therapies, survival rates have
`
`

`

`WO 2008/042236
`
`PCT/US2007/020889
`
`increased little over the last 15 years (Spiro et al., 2002, Am. J. Respir. Crit, Care Med.
`
`166:1166-96).
`
`[01 1]
`
`These four cancers, as well as many others, present as solid tumors that are
`
`composed of heterogeneous cell populations. For example, breast cancers are a mixture of
`
`cancer cells and normal cells, including mesenchymal (stromal) cells, inflammatory cells, and
`
`endothelial cells. Several models of cancer provide different explanations for the presence of
`
`this heterogeneity. One model, the classic model of cancer, holds that phenotypically distinct
`
`cancer cell populations all have the capacity to proliferate and give rise to a new tumor. In
`
`the classical model, tumor cell heterogeneity results from environmental factors as well as
`
`ongoing mutations within cancer cells resulting in a diverse population, of tumon'genic cells.
`
`This model rests on the idea that all populations of tumor cells have some degree of
`
`tumorigenic potential. (Pandis et al., 1998, Genes, Chromosomes & Cancer 122122-129;
`
`Kuukasjrvi et al., 1997, Cancer Res. 57:1597-1604; Bonsing et al., 1993, Cancer 71 :382-391;
`
`Bonsing et al., 2000, Genes Chromosomes & Cancer 82: 173—183; Beerman H et al., 1991,
`
`.Cytometry 122147—54; Aubele M & Werner M, 1999, Analyt. Cell. Path. 19:53; Shen L et al.,
`
`2000, Cancer Res. 60:3884).
`
`[012]
`
`An alternative model for the observed solid tumor cell heterogeneity derives
`
`from the impact of stem cells on tumor development. According to this model, cancer arises
`
`from dysregulation of the mechanisms that control normal tissue development and
`
`maintenance.
`
`(B eachy et al., 2004, Nature 432:324). During normal animal development,
`
`cells of most or all tissues are derived from normal precursors, called stem cells (Morrison et
`
`al., 1997, Cell 882287-98; Morrison et al., 1997, Curr. Opin. Immunol. 9:216—21; Morrison et
`
`al., 1995, Annu. Rev. Cell. Dev. Biol. 11:35-71). Stem cells are cells that: (1) have extensive
`
`proliferative capacity; 2) are capable of asymmetric cell division to generate one or more
`
`kinds of progeny with reduced proliferative and/or developmental potential; and (3) are
`
`

`

`WO 2008/042236
`
`PCT/US2007/020889
`
`capable of symmetric cell divisions for self-renewal or self—maintenance. The best—studied
`
`example of adult cell renewal by the differentiation of stem cells is the hematopoietic system
`
`where developmentally immature precursors (hematopoietic stem and progenitor cells)
`respond to molecular signals to form the varied blood and lymphoid cell types. Other cells,
`
`including cells of the gut, breast ductal system, and skin are constantly replenished from a
`
`small population of stem cells in each tissue, and recent studies suggest that most other adult
`
`tissues also harbor stem cells, including the brain. Tumors derived from a “solid tumor stem
`
`cell” (or “cancer stem cell” from a solid tumor) Subsequently undergoes chaotic development
`through both symmetric and asymmetric rounds of cell divisions. In this stem cell model,
`
`solid tumors contain a distinct and limited (possibly even rare) subset of cells that share the
`
`properties of normal “stem cells”, in that they extensively proliferate and efficiently give rise
`
`both to additional solid‘tumor stem cells (self-renewal) and to the majority of tumor cells of a
`
`solid tumor that lack tumorigenic potential. Indeed, mutations within a long-lived stem cell
`
`population may initiate the formation of cancer stem cells that underlie the growth and
`
`maintenance of tumors and whose presence contributes to the failure of current therapeutic
`
`approaches.
`
`[013]
`
`The stem cell nature of cancer was first revealed in the blood cancer, acute
`
`myeloid leukemia (AML) (Lapidot et al., 1994, Nature 17:645-8). More recently it has been
`
`demonstrated that malignant human breast and colon tumors similarly harbor a small, distinct
`
`population of cancer stem cells enriched for the ability to form tumors in immunodeficient
`
`mice. An ESA+, CD44+, CD24-/low, Lin— cell population in breast tumors was found to be
`
`50-fold enriched for tumorigenic cells compared to unfractionated tumor cells (Al-Hajj et al.,
`
`2003, Proc. Nat ’l Acad. Sci. 100:3983-8). Similarly, the ESA+, CD44+ subpopulation in
`
`colorectal tumors was found to uniquely include tumorigenic cells, and the addition of
`
`CD166 to this profile was able to further enrich for colon cancer stem cells (CoCSC)
`
`

`

`WO 2008/042236
`
`PCT/US2007/020889
`
`(Dalerba et a1. 2007 Proc Nat ’l Acad Sci 104:1 0158-63). The ability to prospectively isolate
`
`the tumorigenic cancer cells has permitted investigation of critical biological pathways that
`
`underlie tumorigenicity in these cells, and thus promises the development of better diagnostic
`
`assays and therapeutics for cancer patients. It is toward this purpose that this invention is
`
`directed.
`
`SUMMARY
`
`Provided are antibodies that specificallybind to a human Delta-like ligand 4
`[014]
`(DLL4) epitope formed by a combination of the human DLL4 N—terminal region (SEQ ID
`
`NO: 27) and human DSL domain (SEQ ID NO: 26), wherein the antibody affects tumor
`
`growth. Also provided is a pharmaceutical composition comprising an antibody of the
`
`present disclosure and a pharmaceutically acceptable vehicle. Further provided is a method
`
`of treating cancer comprising administering a therapeutically effective amount of a DLL4
`
`antibody of the present disclosure.
`
`[015] Additional objects and advantages of the invention will be set forth in part in
`the description which follows, and in part will be obvious from the description, or maybe
`
`learned by practice of the invention. The objects and advantages of the invention will be
`
`realized and attained by means of the elements and combinations particularly pointed out in
`
`the appended claims. It is to be understood that both the foregoing general description and
`
`the following detailed description are exemplary and explanatory only and are not restrictive
`
`of the invention, as claimed. The accompanying drawings, which are incorporated in and
`
`constitute a part of this specification, illustrate several embodiments of the invention and,
`
`together with the description, serve to explain the principles of the invention. In the
`specification and the appended claims, the singular forms “a,” “an,” and “the” include plural
`
`reference unless the context clearly dictates otherwise.
`
`

`

`WO 2008/042236
`
`PCT/US2007/020889
`
`BRIEF DESCRIPTION OF THE DRAWINGS
`
`[016]
`
`Figure 1: Specific Binding of anti—DLL4 21M18 Antibodies to Native Cell-
`
`Surface DLL4 Protein. HEK 293 cells co-transfected with fiill—length DLL4 and GFP were
`
`incubated with anti-DLL4 antibodies and sorted by FACS. Anti-DLL4 antibodies 21Ml4
`
`and 21M18 show specific binding to cells expressing DLL4 as revealed by the linear
`
`relationship between DLL4 antibody binding and GFP expression.
`
`[017]
`
`Figure 2: DLL4 Antibodies Block the Interaction of Human DLL4 with the
`
`Notch Receptor. A) HEK 293 cells expressing DLL4 were incubated with Notch-Fe or
`
`control Fc protein in the presence ofDLL4 or control antibodies. High fluorescence intensity
`
`indicates the presence of Notch and DLL4 binding in the presence of a control antibody (line
`
`2) and 21M12 anti-DLL4‘antibodies (line 5). Low fluorescence intensity indicates the
`
`absence of Notch and DLL4 interactions in the absence of Notch (line 1) and the disruption
`
`of Notch and DLL4 interactions in the presence of anti-DLL4 antibodies 21M18 (line 3) and
`
`21M14 (line 4). B) HEK 293 cells expressing Notch] were incubated with either human or
`
`murine DLL4-Fe. Binding was detected by fluorescently labeled anti-Fe and analyzed by
`
`FACS, with high fluorescence intensity indicative of binding between DLL4 and Notchl
`
`expressing cells. 21M18 blocks binding of human DLL4 (gray squares) but not murine
`
`DLL4 (black circles) to the Notch receptor.
`
`[01 8]
`
`Figure 3: Epitope Mapping of Anti—DLL4 Antibodies. A) Fusion proteins
`
`with nested deletions of the extracellular domain of human DLL4 were incubated in an
`
`ELISA assay with 21Ml4 and 21M18 anti—DLL4 antibodies. No binding above background
`
`was detected in the presence of fusion proteins containing between amino acids 1 to 154 (aa
`
`1-96, white bar with black dots; aa 1-154, black bar with White dots). In contrast, binding
`
`was detected between anti-DLL4 antibodies and all fusion proteins containing between amino
`
`‘ acids 1 to 217, including the DSL domain, of DLL4 (aa 1-217, horizontal striped bar; aa 1-
`
`

`

`WO 2008/042236
`
`PCT/US2007/020889
`
`251, diagonal striped bar; aa 1—283, hatched bar; aa 1-323, gray bar with white dots). B)
`
`Western blots show expression of human DLL4 (h-DLL4) C-terrninal deletion proteins and
`
`murine-hurnan DLL4 chimeric fusion proteins (anti-hFc; top). The DLL4 fusion proteins
`
`comprise one or more of domains 1 to 6, where domains 1 and 2 are N—terminal amino acids
`
`1 to 154; domain 3 is the DSL domain from amino acids 155 to 217; and domains 4, 5, and 6
`
`are each an EGF domain as depicted graphically in C. 21M18 antibodies recognize h-DLL4
`
`protein only in the presence of amino acids 1—217 (hDLL4doml-3). In contrast to the human
`
`protein, fusion pr‘oteins comprising murine DLL4 (m—DLL4) amino acids 1-217 (dom1-3) are
`not recognized by 21M1 8 (m~DLL4 dom1-3zh-DLL4dom4-6). Yet fusion proteins
`V
`
`comprising h-DLL4 amino acids 1-154 (doml—Z) in the presence of murine dom3 are
`recognized by 21M18 (h-DLL4 doml -2:mDLL4don13-6). C) A schematic summary of the
`
`binding data of B is shown. The domain strucmre of DLL4 is shown at top with the DLL4
`
`fusion proteins listed and shown schematically on the lefi side with human protein
`
`represented by light gray and mouse protein represented by dark gray. 21M18 binding to
`
`each DLL4 fragment is indicated by a “+” versus a “—”. D) ELISA analysis of 21M18
`
`binding to DLL4 protein fragments containing substitution of corresponding murine residues
`
`for human residues at select positions. 21M] 8 displays impaired binding to DLL4 protein
`fragments with substitutions at amino acids 68, 69, and 71 (replacement of valine, valine,
`
`and proline) or at amino acids 142 and 144 (replacement of lysine and alanine). E) ELISA
`
`analysis of the binding of antibodies 21M18 and 21M21 to DLL4 protein fragments
`
`containing substitution of corresponding murine residues for human residues at select
`
`positions within the DSL domain. Antibody 21M21 displays impaired binding to human
`
`DLL4 protein fi-agments containing amino acid substitutions at amino acids 161 and 162
`
`(replacement of threonine and serine). As 21M21 does not impair DLL4 function in
`
`

`

`WO 2008/042236
`
`PCT/US2007/020889
`
`signaling assays (see Figure 6), this demonstrates that not all antibodies that bind to the DSL
`
`region impact DLL4 function.
`
`[019]
`
`1 Figure 4: Sequence alignment of the heavy chain variable region. A)
`
`Parental murine 21M1 8 antibody sequence (m-21M18-Vh, top) human expressed framework
`sequence (h-EST-framework, middle) and the humanized 21M18 heavy chain variable region
`
`sequence (21M18- H7, bottom) are shown with conserved amino acid residues shaded in
`
`black. The three CDRS are marked showing retention of parental murine sequences in the
`
`humanized 21M18 antibody. The cysteine residue at Kabat position 52a in CDR2 has been
`changed to a serine and a valine residue without loss of specific binding to D114 in 21M18 H7
`
`and 21M18 H9, respectively. Substitutions within the framework region shown in 4A are
`
`numbered 1-6 with corresponding Kabat positions in the V11 chain 16, 20, 27, 28, 38, 48. B)
`Parental murine 21M1 8 antibody sequence (m—21M18-Vh, top), human germline Vh
`
`sequence (h-germline-Vh, middle), and the humanized 21M18 heavy chain variable region
`
`sequence (21M18- H2, bottom) are shown with conserved amino acid residues shaded in
`
`black. The three CDRS are marked showing retention of parental murine sequences in the
`
`humanized 21 M1 8 antibody. The cysteine residue at Kabat position 52a in CDR2 has been
`changed to a serine and a valine residue without loss ofspecific binding to D114 in 21M18 H7
`
`and 21M18 H9, respectively. The five retained murine residues within the variable
`
`framework region of all heavy chain variants are numbered 1-5 at their corresponding Kabat
`
`positions 20, 28, 38, 48, and 69.
`
`[020]
`
`' Figure 5: Sequence alignment of the light chain variable region. Parental
`
`murine 21M18 antibody sequence (m-21M18-Vk, top), human germline sequence (h-
`
`germline Vk, bottom), and humanized 21M18 light chain variable region sequence (21M18-
`
`L2, middle) are shown with conserved amino acid residues shaded in black. The three CDRS
`
`are marked showing retention of parental mun'ne sequences in the humanized 21M18
`
`10
`
`

`

`WO 2008/042236
`
`PCT/US2007/020889
`
`antibody. The two retained murine residues within the variable framework region are
`
`numbered 1-2 at their corresponding Kabat positions 22 and 36-
`
`[021]
`
`Figure 6: DLL4 Antibodies Block Notch Signaling. HeLa cells co-
`
`transfected with Hesl-Luc reporter and Renilla luciferase reporter vectors were incubated
`
`. with DLL4-Fe protein in the presence or absence of anti-DLL4 antibodies. Decreased
`
`luciferase levels demonstrate loss of DLL4 Notch pathway activation by 21M14 and 21M18
`
`antibodies.
`
`[022]
`Figure 7: DLL4 Antibodies Modulate Expression of Notch Target Genes in
`Colon Tumors. A) C8 colon tumors treated with anti-DLL4 21M18 antibodies or PBS
`
`(Control) were isolated and expression of HESl and ATOH-l determined by quantitative RT-
`
`PCR. Relative gene expression (y-axis) compared to control treated cells shows that
`
`treatment with anti—DLL4 antibodies decreased expression of HESI and increased expression
`
`of ATOH—l . B) Relative expression ratio (y—axis) of HES 1 versus ATOHI in mouse lineage—
`
`depleted OMP-Cl 1 colon tumor cell colonies is shown. C11 colonies overlaid with 3T3 cells
`
`overexpressing DLL4 (3T3+DLL4) showed an increased in the HESl/ATOHI expression
`
`ratio compared to colon cells overlaid with 3T3 cells (3T3) or not exposed to cell overlay
`
`(Control). This increase in the HES l/ATOI-Il expression ratio was eliminated by incubation
`
`with 10 ug/mL 21M18 antibodies (21M18) or 5 uM-secretase inhibitor DBZ (5 uM GSI).
`
`[023]
`
`Figure 8: DLL4 Antibodies Reduce Tumor Growth. NOD/SCID mice were
`
`injected with dissociated UM-C4 cells and treated with anti-DLL4 21M18 antibodies (n=5) or
`
`PBS (n=10). Treatment with 21M1 8 antibodies (diamonds) reduced tumor growth starting on
`
`day 23, and up to 54% reduction was observed by day 48 compared to PBS injected controls
`
`(black squares).
`
`[024]
`
`Figure 9: Treatment with DLL4 Antibodies Reduces the Number of
`
`Proliferating Tumor Cells in Vivo. C8 Colon tumors treated with anti—DLL4 21Ml 8
`
`11
`
`

`

`WO 2008/042236
`
`PCT/U52007/020889
`
`antibodies or control Ab were isolated. Immunocytochemistry with an antibody against Ki67
`
`showed a reduction in the number of proliferating cells in 21M18 treated tumors compared to
`
`control.
`
`[025]
`
`Figure 10: Treatment with DLL4 Antibodies in Combination with
`
`Fluorouracil (5 -FU) Reduces Tumor Growth. NOD/SCH) mice were injected with
`
`dissociated UM-C4 cells and treated with anti—DLL4 antibodies or PBS in the presence or
`
`absence of S—FU. A) Treatment with 21M18 antibodies in combination with S—FU (circles,
`
`dashed line) reduced tumor growth 46 days after injection of tumor cells to a greater degree
`than treatment with either S—FU (triangles, solid line) or 21'M1 8 antibodies (diamonds, dotted
`
`line) alone and to a greater degree than PBS injected controls (squares, solid line). Tumor
`
`volume in mm3 is indicated on the y—axis. B) Plots of tumor measurements on day 46 from
`
`individual animals. Each dot represents one animal. Treatment with 21M18 antibodies or 5-
`
`FU each reduced tumor size (mm3) compared to control. Furthermore, combination treatment
`
`with 21M18 antibodies and S-FU had an additive effect,» reducing tumor size to 1/5 the size of
`
`control.
`
`[026]
`
`Figure 1 1: Treatment with DLL4 Antibodies in Combination with anti—
`
`EGFR Antibodies Reduces Tumor Growth. NOD/SCH) mice were injected with dissociated
`
`UM-C4 cells and treated with anti-DLL4 antibodies or PBS in the presence or absence of
`
`anti-EGFR antibodies. Plots of tumor measurements on day 46 fiom individual animals are
`
`shown. Each dot represents one animal. Treatment with 21M18 antibodies or anti—EGFR
`
`antibodies each reduced tumor size (mm3) compared to control. Furthermore, combination
`
`treatment with 21M18 and anti-EGFR antibodies had an additive effect, reducing tumor size
`
`to less than 1/5 the size of control.
`
`[027]
`
`Figure 12: Anti-DLL4 mAb 21M18 and Irinotecan Act Synergistically to
`
`Inhibit Colon Tumor Growth. NOD/SCID mice were injected with dissociated C8 cells and
`
`12
`
`

`

`WO 2008/042236
`
`PCT/US2007/020889
`
`treated with anti-DLL4 antibodies or control antibody in the presence or absence of
`
`Irinotecan. A) Treatment with murine 21M1 8 antibodies (circles) or Irinotecan (triangles)
`alone each reduced tumor volume (y-axis mm3) compared to control treated animals (black
`
`squares). However, combination treatment with 21M18 and Irinotecan (inverse triangles)
`
`had a synergistic effect, completely eliminating tumor growth for up to 55 days post-cell
`
`injection. B) Treatment with humanized 21M18 (h21Ml 8) in combination with Irinotecan
`
`(irtcn) (circles) has similar efficacy as murine 21M18 (m21M18) (triangles) as compared to
`
`control antibody (black squares) or control antibody with Irinotecan (triangles).
`[028’]
`Figure 13: Combination Anti-DLL4 21M18 and Irinotecan Treatment
`
`’Prevents Colon Tumor Re-Growth. NOD/SCID mice were injected with dissociated C8 cells
`
`and treated with Irinotecan or Irinotecan in combination with anti-DLL4 21M 18 antibodies
`
`(n=lO per group). A) Treatment with Irinotecan alone slowed colon tumor growth, but
`
`growth continued after cessation of treatment on day 56 (* arrow) in all but two treated
`
`animals. B) In contrast, treatment with a combination of Irinotecan and anti—DLL4 21M18
`
`antibodies eliminated colon tumor growth both during treatment and for up to five weeks
`
`following cessation of treatment on day 56 in all ten treated animals. Each line represents the
`
`growth curve for an individual animal.
`
`[029]
`
`Figure 14: Combination Anti-DLL4 21M18 and Irinotecan Treatment
`
`Inhibits the Growth of Established Colon Tumors More Effectively than Single Therapy
`
`Treatment. NOD/SCID mice were injected with dissociated C8 cells and treated with anti-
`
`DLL4 antibodies or control antibody in the presence or absence oflrinotecan. Treatment
`
`with 21M18 antibodies (diamonds) or Irinotecan (triangles) alone each reduced tumor
`
`volume (y—axis mm3) compared to control treated animals (black squares). However,
`
`combination treatment with 21M18 plus Irinotecan (inverse triangles) inhibited tumor growth
`
`more effectively than either 21M18 or Irinotecan treatment alone.
`
`13
`
`

`

`WO 2008/042236
`
`PCT/US2007/020889
`
`[030]
`
`Figure 15: Tumors Treated with Anti-DLL4 Antibodies Show Decreased
`
`Numbers of Tumorigenic Cells. Immunocompromised mice (n = 10 per group) were injected
`
`with decreasing dosages of tumor cells from the experiment shown in Figure 14 that had been
`
`treated with either control antibody, Irinotecan plus control antibody, DLL4 :2le
`antibodies alone, or a combination ofDLL4 21Ml 8 antibodies and Irinotecan (Combination).
`
`A) Results of tumor take rates on day 81. Tumor volume (mm3) was graphed compared to
`
`the number of human tumor cells injected: 900, 300, 100, and 50 for each treatment group.
`
`The'number of animals with detectable tumors over the teri injected animals for each tumor
`
`cell dose is recorded below the graph of tumor volume for each cell dose with control treated
`
`tumor cells on the left (filled circles), anti-DLL4 21M18 antibody treated tumor cells second
`
`to the left (open squares), Irinotecan treated tumor cells second to the right (filled triangles),
`
`and Combination treated tumor cells on the right (open circles). B) The stem cell frequency
`
`on day 81 was calculated. The proportion of cancer stem cells (y-axis) from control treated
`
`(left) compared to anti-DLL4 treated (second from left), Irinotecan only treated (second from
`right), and Combination treated (right) tumor cells is plotted with the 95% confidence
`
`interval. The anti-DLL4 treated group has a statistically significant difference versus the
`
`control group (*) and the combination group is significantly different versus both the control
`
`(*) and Irinotecan alone groups(**).
`
`[031]
`
`Figure 16: Anti-DLL4 21M18 and Irinotecan Combination Treatment Delays
`
`Tumor Recurrence. Immunocompromised mice were injected with dissociated C8 cells and
`
`established tumors of approximately 150 mm
`
`3 were treated with a combination of Irinotecan
`
`(45 mg/kg, dosed twice a week) with either anti-DLL4 21M18 antibodies or control
`
`antibodies for 32 days after which Irinotecan treatment was halted. Treatment with the either
`
`the control antibody or 21M18 continued. Reoccurrence of tumors by tumor volume (y-axis)
`
`was delayed in 21M18 treated animals (triangles) as compared to controls (circles).
`
`