Development of Therapeutic Antibodies Targeting Human Acute Myeloid Leukemia Stem Cells

Funding Type: 
Disease Team Research I
Grant Number: 
Award Value: 
Disease Focus: 
Blood Cancer
Solid Tumor
Collaborative Funder: 
Stem Cell Use: 
Cancer Stem Cell
Cell Line Generation: 
Cancer Stem Cell
Public Abstract: 

Acute myeloid leukemia (AML) is a cancer of the blood and bone marrow that is rapidly fatal within months if untreated. Even with aggressive treatment, including chemotherapy and bone marrow transplantation, five-year overall survival rates range between 30-40%. Evidence indicates that not all cells in this cancer are the same, and that there is a rare population of leukemia stem cells (LSC) that are responsible for maintaining the disease. Thus, in order to cure this cancer, all LSC must be eliminated, while at the same time sparing the normal blood forming stem cells in the bone marrow. We propose to develop therapeutic antibodies directed against surface markers present in much larger amounts on LSC than on the surface of normal blood forming stem cells. We recently identified and validated several such protein markers including CD47, which we determined contributes to leukemia development by blocking the ingestion and removal of leukemia cells by immune system cells called macrophages. In this way, CD47 acts as a “don’t eat me” signal on LSC. Moreover, we determined that monoclonal antibodies (mAbs) directed against CD47, able to block its interaction with macrophages, mask the “don’t eat me” signal resulting in ingestion and elimination of leukemia in mouse pre-clinical models. We propose a combination of clinical studies, basic research, and pre-clinical development to prepare a therapeutic antibody directed against CD47 and/or other LSC-specific proteins for Initial New Drug (IND) filing with the FDA, and then a Phase I clinical trial to be conducted at {REDACTED} and in the Collaborative Funding Partner country. In collaboration with the pioneering Collaborative Funding Partner country AML Working Group, we will track expression of the LSC proteins in patient samples and correlate with clinical outcomes. This will allow us to identify particular LSC proteins that must be targeted to achieve cure, thereby prioritizing candidate therapeutic antibodies for clinical development. Concurrently, we will conduct basic research and pre-clinical development to prepare these candidates. Basic research during years 1 and 2 will focus on the characterization of anti-CD47 mAb efficacy, investigation of mAb targeting of additional LSC molecules, and determination of efficacy in combinations with anti-CD47. Pre-clinical development during years 1 and 2 will focus on blocking anti-CD47 mAbs, including antibody humanization and large animal model pharmacologic and toxicity studies. Similar studies will be conducted with the most promising antibodies resulting from our basic research. During years 3-4, we will proceed with GMP grade production of the best candidate, followed by efficacy testing in mouse models and large animal models. Finally, in year 4, we will prepare an IND filing with the FDA/MHRA and develop a Phase I clinical trial with this antibody for the treatment of AML. Ultimately, therapeutic antibodies specifically targeting AML LSC offer the possibility of less toxicity with the potential for cure.

Statement of Benefit to California: 

Acute myeloid leukemia (AML) is an aggressive malignancy of the bone marrow with nearly 13,000 new diagnoses annually in the US and 2,200 in the Collaborative Funding Partner country. Current standard of care for medically fit patients consists of several cycles of high dose chemotherapy, and often includes allogeneic hematopoietic cell transplantation. Even with these aggressive treatments, which cause significant morbidity and mortality, relapse is common and the five-year overall survival is 30-40%, but <10% in patients with relapsed or refractory disease or in the majority of AML patients who are over age 65. The goal of this research proposal is to prepare therapeutic antibodies directed against AML stem cell-specific antigens for IND filing with the FDA and a Phase I clinical trial. There are several potential benefits of this research for California: (1) most importantly, this research has the potential to revolutionize current clinical practice and provide a targeted therapy for AML that offers the possibility of less toxicity with the potential for cure; (2) this research will directly contribute to the California economy by funding a contract manufacturing organization to generate and produce GMP-grade clinical antibody, by employing several individuals who will be essential for the conduct of these studies, and through the purchase of equipment and reagents from California vendors; (3) additional clinical and economic benefits for California will derive from the potential application of clinical agents developed here to a number of other human cancers and cancer stem cells; (4) our animal models indicate that a significant fraction of patients with fatal AML can be cured, resulting in savings on their clinical care plus their return as productive contributors to the California economy; (5) if our therapeutic antibodies show clinical benefit in AML, they will be commercialized, and under CIRM policy, profits derived from treating insured patients and lower cost therapies for uninsured patients, would enrich the state and the lives of its citizens; (6) finally, this research has the potential to maintain California as the national and world-wide leader in stem cell technology.

Progress Report: 

Our program is focused on producing new therapeutic candidates to prolong remission and potentially cure highly lethal cancers where patients have few alternative treatment options. We have selected Acute Myelogenous Leukemia (AML) as the initial clinical indication for evaluating our novel therapeutics, but anticipate a full development program encompassing many other types of solid tumor cancers.

Our strategy is to develop an antibody that binds to and eliminates the cancer-forming stem cells in leukemia and other solid tumors. While current cancer treatments (e.g. surgery, chemotherapy, radiation) will frequently get rid of the bulk of the tumor, they rarely touch the tiny number of cancer stem cells that actually re-generate the masses of cancer cells that have been eliminated. When the latter occurs, the patient is described as having a relapse, leading to a disease recurrence with poor prognosis. Our strategy is to eliminate the small number of cancer-regenerating stem cells by targeting cell membrane proteins expressed by these cells.

We have discovered that many cancer cells coat themselves with a protein called CD47 that prevents them from being eaten and disposed of by the patient’s blood cells. In this context, CD47 can be considered a ‘don’t eat me’ signal that protects the cancer cells from being phagocytosed i.e. ‘eaten’. The antibody we are developing binds to and covers the ‘don’t eat me’ CD47 protein, so that the patient’s blood cells are now able to ‘eat’ the cancer cells by standard physiological responses, and eliminate them from the body.

Developing an antibody such as this for use in humans requires many steps to evaluate it is safe, while at the same ensuring it targets and eliminates the cancer forming stem cells. The antibody must also ‘look’ like a human antibody, or else the patient will ‘see’ it as a foreign protein and reject it. To achieve these criteria, we have made humanized antibodies that bind to human CD47. We have shown that the antibodies eliminate cancer cells in two ways: (i) blood cells from healthy humans rapidly “ate” and killed leukemia cells collected from separate cancer patients when the anti-human CD47 antibody was added to a mixture of both cell types in a research laboratory test tube; (ii) the anti-human CD47 antibody eliminates human leukemia cells collected from patients, then transferred into special immunodeficient mice which are unable to eliminate the human tumor cells themselves. In these experiments, the treated mice remained free of the human leukemia cells for many weeks post-treatment, and could be regarded as being cured of malignancy.

To show the antibodies were safe, we administered to regular mice large amounts of a comparable anti-mouse CD47 antibody on a daily basis for a period of many months. No adverse effects were noted. Unfortunately our antibody to human CD47 did not bind to mouse CD47, so it’s safety could not be evaluated directly in mice. Since the anti-human CD47 antibody does bind to non-human primate CD47, safety studies for our candidate therapeutic need to be conducted in non-human primates. These studies have been initiated and are in progress. Following administration of the anti-human CD47 antibodies, the non-human primates will be monitored for clinical blood pathology, which, as in humans, provides information about major organ function as well as blood cell function in these animals.

The next step after identifying an antibody with strong anti-cancer activity, but one that can be safely administered to non-human primates without causing any toxic effects, is to make large amounts of the antibody for use in humans. Any therapeutic candidate that will be administered to humans must be made according to highly regulated procedures that produce an agent that is extremely “clean”, meaning free of viruses, other infectious agents, bacterial products, and other contaminating proteins. This type of production work can only be performed in special facilities that have the equipment and experience for this type of clinical manufacturing. We have contracted such an organization to manufacture clinical grade anti-human CD47 antibodies. This organization has commenced the lengthy process of making anti-CD47 antibody that can be administered to humans with cancer. It will take another 18 months to complete the process of manufacturing clinical grade material in sufficient quantities to run a Phase I clinical trial in patients with Acute Myelogenous Leukemia.

Our program is focused on producing new therapeutic candidates to prolong remission and potentially cure highly lethal cancers where patients have few alternative treatment options. Our strategy is to develop an antibody that will eliminate the cancer stem cells which are the source of the disease, and responsible for the disease recurrence that can occur months-to-years following the remission achieved with initial clinical treatment. The cancer stem cells are a small proportion of the total cancer cell burden, and they appear to be resistant to the standard treatments of chemotherapy and radiation therapy. Therefore new therapeutic approaches are needed to eliminate them.

In year 2 of the CIRM award, we have continued to develop a clinical-grade antibody that will eliminate the cancer stem cells in Acute Myelogenous Leukemia (AML). We have identified several antibodies that cause human leukemia cells to be eaten and destroyed by healthy human white blood cells when tested in cell culture experiments. These antibodies bind to a protein called CD47 that is present on the outer surface of human leukemia cells. The anti-CD47 antibodies can eliminate leukemia growing in mice injected with AML cells obtained from patients. We have now extensively characterized the properties of our panel of anti-CD47 antibodies, and have identified the lead candidate to progress though the process of drug development. There are several steps in this process, which takes 18-24 months to fully execute. In the last 12 months, we have focused on the following steps:

(i) ‘Humanization’ of the antibody: The antibody needs to be optimized so that it looks like a normal human protein that the patient’s immune system will not eliminate because it appears ‘foreign’ to them.
(ii) Large scale production of the antibody: To make sufficient quantities of the antibody to complete the culture and animal model experiments required to progress to clinical safety trials with patients, we have contracted with a highly experienced manufacturing facility capable of such large-scale production. We have successfully transferred our antibody to them, and they have inserted it into a proprietary expression cell that will produce large amounts of the protein. This process is managed through weekly interactions with this contract lab. They send us small amounts of the material from each step of their manufacturing process and we test it in our models to ensure the antibody they are preparing retains its anti-cancer properties throughout production.
(iii) Pre-clinical safety studies: The antibody must be tested extensively in animals to ensure it does not cause serious limiting damage to any of the normal healthy tissues in the recipient. We have spent much of the last 12 months performing these types of safety experiments. The antibody has been administered to both mice and non-human primates and we have evaluated their overall health status, as well as analyzing their blood cells, blood enzyme levels, and urine, for up to 28 days. We have also collected samples of their organs and tissues to evaluate for abnormalities. Thus far, these assessments have appeared normal except for the development of a mild anemia a few days after the initial antibody injection. Subsequent experiments indicate that this anemia can be managed with existing approved clinical strategies
(iv) Determination of optimal dose: We have used mice injected with human cancer cells from AML patients, and determined how much antibody must be injected into these mice to produce a blood level that destroys the leukemia cells. This relationship between antibody dose and anti-cancer activity in the mouse cancer model enables us to estimate the dose to administer to patients.

Hematologic tumors and many solid tumors are propagated by a subset of cells called cancer stem cells. These cells appear to be resistant to the standard cancer treatments of chemotherapy and radiation therapy, and therefore new therapeutic approaches are needed to eliminate them. We have developed a monoclonal antibody (anti-CD47 antibody) that recognizes and causes elimination of these cancer stem cells and other cells in the cancer, but not normal blood-forming stem cells or blood cells. Cancer stem cells regularly produce a cell surface ‘invisibility cloak’ called CD47, a ‘don’t eat me signal’ for cells of the native immune system. Anti-CD47 antibody counters the ‘cloak, allowing the patient’s natural immune system eating cells, called macrophages, to eliminate the cancer stem cells.
As discussed in our two-year report, we optimized our anti-CD47 antibody so that it looks like a normal human protein that the patient’s immune system will not eliminate because it appears ‘foreign’. In this third year of the grant, we initiated the pre-clinical development of this humanized antibody, and assigned the antibody the development name of Hu5F9. Our major accomplishments in the third year of our grant are as follows:
(i) In addition to the hematological malignancies we have studied in previous years, we have now demonstrated the Hu5F9 is effective at inhibiting the growth and spread throughout the body [metastasis] of a large panel of human solid tumors, including breast, bladder, colon, ovarian, glioblastoma [a very aggressive brain cancer], leiomyosarcoma, head & neck squamous cell carcinoma, and multiple myeloma.
(ii) We have performed extensive studies optimizing the production and purification of Hu5F9 to standards compatible with use in humans, including that it is sterile, free of contaminating viruses, microorganisms, and bacterial products. We will commence manufacturing of Hu5F under highly regulated sterile conditions to produce what is known as GMP material, suitable for use in humans.
(iii) Another step to show Hu5F9 is safe to administer to humans is to administer it to experimental animals and observe its effects. We have demonstrated that Hu5F9 is safe and well tolerated when administered to experimental animals. Notably, no major abnormalities are detected when blood levels of the drug are maintained in the potentially therapeutic range for an extended duration of time.
(iv) We have initiated discussions with the FDA regarding the readiness of our program for initiating clinical trials, which we anticipate to start in the first quarter of 2014. To prepare for these trials we have established a collaboration between the Stanford Cancer Institute and the University of Oxford in the United Kingdom, currently our partners in this CIRM-funded program.
To our knowledge, CD47 is the first common target in all human cancers, one which has a known function that enables cancers to grow and spread, and one which we have successfully targeted for cancer therapy. Our studies show that Hu5F9 is a first-in-class therapeutic candidate that offers cancer treatment a totally new mechanism of enabling the patient’s immune system to remove cancer stem cells and their metastases.

Hematologic tumors and many solid tumors are driven by a subset of cells called cancer stem cells. These cancer stem cells must be eliminated for cures, however, they have been found to be resistant to the standard cancer treatments of chemotherapy and radiation therapy. Therefore, new therapeutic approaches are needed to target these abnormal stem cells. Previously, we found that cancer stem cells have developed a clever way to hide from the patient’s immune system. They display a protein called CD47 on their surface that signals to the immune system “don’t eat me”, thereby preventing their elimination. We have developed a monoclonal antibody (anti-CD47 antibody) that blocks this signal leading to elimination of these cancer stem cells, but not normal most normal cells, by the natural immune system. In our pre-clinical studies, we showed that anti-CD47 antibodies eliminates cancer cells and cancer stem cells from many different types of human cancer including: leukemia, breast cancer, colon cancer, prostate cancer, ovarian cancer, and others. In addition, anti-CD47 antibodies are effective at preventing and even eliminating metastases in animal models. These results indicate that anti-CD47 antibodies have great potential for the treatment of human cancer.

In order to develop this approach into a clinical therapeutic, we first optimized our anti-CD47 antibody so that it looks like a normal human protein that the patient’s immune system will not reject. Over the course of this grant project, we have conducted the pre-clinical development of this humanized antibody, termed Hu5F9-G4.

(1) Hu5F9-G4 has been manufactured according to Good Manufacturing Practices (GMP) as required by the United States Food and Drug Administration (FDA) for administration to humans. The drug product was manufactured and tested to be free of contaminants and is now ready for clinical use.

(2) Hu5F9-G4 has undergone extensive testing to investigate potential toxic effects in humans. According to FDA regulatory guidelines, Hu5F9-G4 was tested in experimental animals where it was given in various increasing doses. In all studies, Hu5F9-G4 was well-tolerated and caused no serious side effects.

(3) We have developed a phase 1 first-in-human clinical trial protocol for the investigation of Hu5F9-G4 in patients with solid tumors. In addition, we have prepared all the necessary documentation and clinical operations plans necessary to execute this clinical trial.

(4) We have submitted the necessary information on anti-cancer activity, manufacturing, safety, and clinical trial plans to the FDA in an Investigational New Drug (IND) application. This application was approved by FDA for the clinical trial in patients with solid tumors.

(5) We continue to develop parallel clinical trial plans for a phase 1 study in patients with acute myeloid leukemia (AML), and anticipate submitting our regulatory filing in 2015.
In summary, our studies show that Hu5F9-G4 is a first-in-class therapeutic candidate that offers cancer treatment through a totally new mechanism of enabling the patient’s immune system to remove cancer stem cells and prevent their metastases.