Predictive biomarker for cancer therapy

Abstract

The present invention relates generally to the identification of patients suffering from cancer whether they will respond to specific therapies. More particularly the invention relates to a method and means for identifying responder to a therapy TLR-9 agonists.

Claims

1. A method for treating a patient suffering from cancer or an autoimmune disease with TLR-9 agonist MGN1703, wherein said method comprises: taking a blood sample from a patient suffering from cancer or an autoimmune disease; determining the frequency of activated natural killer T (NKT) cells of the blood sample of the patient; predicting or monitoring whether the patient will respond or responds to the treatment with the TLR-9 agonist MGN1703, by evaluating whether the patient has a frequency of at least 3% of activated NKT cells of the whole NKT cell population; and administering, or continuing to administer, the TLR-9 agonist MGN1703 to a patient having a frequency of at least 3% of activated NKT cells of the whole NKT cell population.

2. The method of claim 1, wherein previously an induction therapy with a non-DNA drug was performed on the patient.

3. The method of claim 1, wherein said patient suffers from cancer.

4. The method of claim 3, wherein the TLR-9 agonist MGN1703 is part of a pharmaceutical composition.

5. The method of claim 4, wherein the pharmaceutical composition is a vaccine.

6. The method of claim 1, wherein the step of determining the frequency of activated natural killer T (NKT) cells of a blood sample of the patient comprises the steps of: staining the blood sample with fluorescence-labelled antibodies: Anti CD3-FITC, Anti CD56-PE, Anti CD69-APC; incubating the sample; and performing fluorescence activated cell sorting (FACS), wherein the activated NKT cells are gated as CD3 positive, CD56 positive and CD69 positive cells.

Description

BRIEF SUMMARY OF THE FIGURES

(1) FIG. 1: ROC curves in verum patients for activated NKT cells.

(2) FIG. 2: Activated NKT cells versus sensitivity and specificity.

(3) FIG. 3: a: Kaplan-Meier plot in biomarker positive patients for verum and placebo; b: Kaplan-Meier curve for the biomarker negative patients in both the verum and placebo arm.

(4) FIG. 4: a: Kaplan-Meier plot of biomarker positive patients versus biomarker negative patients in the verum arm; b: Kaplan-Meier plot of the placebo group for biomarker positive and negative patients.

DETAILED DESCRIPTION OF THE INVENTION

(5) Within the meaning of the present invention MGN1703 designates a DNA construct of a covalently closed partially self-complementary DNA chain having a double stranded stem and single stranded terminal loops bearing unmethylated CG motifs.

(6) A stem according to the present disclosure shall be understood as a DNA double strand formed by base pairing either within the same DNA molecule (which is then partially self-complementary) or within different DNA molecules (which are partially or completely complementary). Intramolecular base pairing designates base pairing within the same molecules, and base pairing between different DNA molecules is termed as intermolecular base-pairing.

(7) A loop within the meaning of the present disclosure shall be understood as an unpaired, single-stranded region either within or at the end of a stem structure. A hairpin is a distinct combination of a stem and a loop, which occurs when two self-complementary regions of the same DNA molecule hybridize to form a stem with an unpaired loop.

(8) A dumbbell-shape describes a linear DNA construct with hairpins at both ends flanking a stem region. Thus, a linear DNA construct within the context of the present disclosure describes a linear dumbbell-shaped DNA construct comprising single stranded loops at both ends of a double stranded DNA stem.

(9) Immunomodulation according to the present disclosure refers to immune activation and immunosuppression. Immune activation means preferentially that effector cells of the immune system are activated in order to proliferate, migrate, differentiate or become active in any other form. B cell proliferation for instance can be induced without co-stimulatory signals by immune activating DNA molecules, which normally require a co-stimulatory signal from helper T-cells.

(10) Immunosuppression on the other hand shall be understood as reducing the activation or efficacy of the immune system. Immunosuppression is generally deliberately induced to prevent for instance the rejection of a transplanted organ, to treat graft-versus-host disease after a bone marrow transplant, or for the treatment of autoimmune diseases such as, for example, rheumatoid arthritis or Crohn's disease.

(11) In this context, immunomodulation may also refer to influencing the nature or the character of an immune reaction, either by affecting an immune reaction, which is still developing or maturing or by modulating the character of an established immune reaction.

(12) The term vaccination used in this disclosure refers to the administration of antigenic material (a vaccine) to produce immunity to a disease. Vaccines can prevent or ameliorate the effects of infection by many pathogens such as viruses, fungi, protozoan parasites, bacteria but also of allergic diseases and asthma, as well as of tumors. Vaccines typically contain one or more adjuvants, e g immune activating nucleic acids, used to boost the immune response. Vaccination is generally considered to be the most effective and cost-effective method of preventing infectious and other diseases.

(13) The material administered can, for example, be live but weakened forms of pathogens (bacteria or viruses), killed or inactivated forms of these pathogens, purified material such as proteins, nucleic acids encoding antigens, or cells such as tumor cells or dendritic cells. In particular, DNA vaccination has recently been developed. DNA vaccination works by insertion (and expression, triggering immune system recognition) of DNA encoding antigens into human or animal cells. Some cells of the immune system that recognize the proteins expressed will mount an attack against these proteins and against cells expressing them. One advantage of DNA vaccines is that they are very easy to produce and store. In addition, DNA vaccines have a number of advantages over conventional vaccines, including the ability to induce a wider range of immune response types.

(14) Vaccination can be used as a prophylactic approach, leading to immunity against the antigen in the vaccinated, healthy individual upon exposure to the antigen. Alternatively, a therapeutic vaccination can cause an improved response of the immune system of the vaccinated, diseased individual, by guiding the immune system of the individual towards the antigens. Both prophylactic and therapeutic vaccination can be applied to humans as well as animals.

(15) The term cancer comprises cancerous diseases or a tumor being treated or prevented that is selected from the group comprising, but not limited to, mammary carcinomas, melanoma, skin neoplasms, lymphoma, leukemia, gastrointestinal tumors, including colon carcinomas, stomach carcinomas, pancreas carcinomas, colon cancer, small intestine cancer, ovarial carcinomas, cervical carcinomas, lung cancer, prostate cancer, kidney cell carcinomas and/or liver metastases.

(16) Autoimmune diseases according to the present disclosure comprise rheumatoid arthritis, Crohn's disease, systemic lupus (SLE), autoimmune thyroiditis, Hashimoto's thyroiditis, multiple sclerosis, Graves' disease, myasthenia gravis, celiac disease and Addison's disease.

(17) During experiments with immune activating dumbbell-shaped DNA constructs with unmethylated CG sequences in the single-stranded terminal loops it turned out that elevated frequencies of certain cell types of the immune activation and effector pathway were related to a successful therapy with the DNA constructs.

(18) In vertebrates, the so-called toll-like receptors (TLRs) are part of the innate immune system. TLRs are a family of specialized immune receptors that induce protective immune responses when they detect highly conserved pathogen-related molecular patterns, such as proteins, lipid structures, sacharidic structures, and certain nucleic acids. Synthetic agonists for several TLRs, including TLR-3, TLR-4, TLR-7, TLR-8, and TLR-9, have been or are being developed for the treatment of cancer, generally with the intention to activate the immune system in the presence of tumours. TLR-9 recognizes the presence of unmethylated CG-containing DNA sequences, which are typically found in bacteria, but practically never in human genomic DNA. Thus, unmethylated CG-containing DNA sequences have been designed as artificial TLR-9 agonists. The effect of such unmethylated CG-containing DNA constructs depends on their interaction with TLR-9, and DNA-protein interaction depends on the conformation of both DNA and protein. Experimental data demonstrate that dumbbell-shaped DNA molecules are surprisingly suitable for the induction of an immune response.

(19) In order to identify a tool for predicting whether a cancer patient will respond to the application of a TLR-9 agonist, a dumbbell-shape DNA construct having unmethylated CG motifs in the single stranded terminal loops was used.

(20) In total, 46 patients suffering from metastatic colorectal cancer who had previously been treated for 4.5 to 6 month in a standard first-line combination therapy with or without a human monoclonal antibody inhibiting vascular endothelial growth factor A were selected for the study. After a treatment-free interval of ca. 1-6 weeks, a randomization of the patients was performed, so that 32 patients received 60 mg per dose of the DNA construct MGN1703 and 14 patients received a placebo, each patient twice weekly by subcutaneous injection.

(21) After 12 weeks of treatment, all patients were examined for tumour-progression. Based on the presence or absence of tumour progression, patients were divided into two groups, designated as PFS groups. Patients whose tumour had not progressed were designated as progression-free patients and labelled PFS1, whiles patients with tumour progression were labelled as PFS0. Obviously, a lack of tumour progression (PFS1) indicates a possible response to the treatment, while tumour growth (PFS2) indicates a lack of response. Treatment was continued for each patient until tumour progression was found.

(22) Table 1 summarizes the results. It is obvious that nearly all progression-free patients received the DNA construct.

(23) TABLE-US-00001 TABLE 1 Designation of patients Group designation Treatment arms PFS0 PFS1 Total MGN1703 19 13 32 Placebo 13 1 14 Total 32 14 46

(24) Prior to the first application of MGN1703 or placebo (at baseline) blood samples of all patients were collected. The distribution of the following immunovariables, e.g. part of whole PBMC population and relation of activated and non-activated sub populations was determined A correlation of the respective PFS group designation after 12 weeks and all immunovariables at baseline was investigated. The following immunovariables were determined: monocytes, activated monocytes 1, activated monocytes 2, B-Cells, activated B-cells, T-cells, activated T-cells, natural killer (NK)-cells, activated NK-cells, NKT-cells, activated NKT-cells, plasmacytoid dendritic cells (pDCs), activated pDCs, myeloid dendritic cells (mDCs), and activated mDCs. Table 2 summarizes the context of cell types, CDs and the determined frequencies.

(25) TABLE-US-00002 TABLE 2 Relation of cell types, CDs and determined frequencies Frequency Cell Type Identified by as percentage of monocytes CD14+ all PBMCs activated monocytes 1 CD14+/CD86+ all monocytes activated monocytes 2 CD14+/CD169+ all monocytes B-cells CD19+ all PBMCs activated B-cells CD19+/CD86+ all B-cells T-cells CD3+/CD56? all PBMCs activated T-cells CD3+/CD56?/CD69+ all T-cells natural killer (NK-)cells CD3?/CD56+ all PBMCs activated NK-cells CD3?/CD56+/CD69+ all natural killer cells NKT cells CD3+/CD56+ all PBMCs activated NKT cells CD3+/CD56+/CD69+ all NKT cells plasmacytoid dendritic cells Lin1?/CD123+/HLA- all PBMCs (pDCs) DR+ activated pDCs Lin1?/CD123+/HLA- all pDCs DR+/CD40+ myeloid dendritic cells Lin1?/CD11c+/HLA- all PBMCs (mDCs) DR+ activated mDCs Lin1?/CD11c+/HLA- all mDCs DR+/CD86+

(26) To assess whether one of the immunovariables may serve as a proper biomarker, a so-called Cox regression was calculated for each immunovariable. A Cox regression allows estimating the effect of parameter(s) without any consideration of the hazard function. The resulting hazard ratio from such a calculation should be below 1 and related to a significant p value being below 0.05. Otherwise the observed effect would not be related to the applied TLR-9 agonist. Those criteria applied only to activated NKT cells having a hazard ratio of about 0.933 and a p value of 0.0309.

(27) Surprisingly, the percentage of activated NKT cells could be used to predict treatment success within the verum group. The relationship between the percentage of activated NKT cells and the PFS group status was studied using advanced, well-established statistical analyses within the verum group.

(28) A receiver operating characteristic (ROC) curve, or simply ROC curve, shows the performance of a binary classifier system as its discrimination threshold is varied. ROC curves are created by plotting the fraction of true positives out of the positives versus the fraction of false positives out of the negatives at various threshold settings. The true positives are also designated as sensitivity and the false positives is one minus the specificity or true negative rate.

(29) ROC analysis is used in medicine, radiology, biometrics, and other areas for many decades and is increasingly used in machine learning and data mining research. In biomarkers, it can be used to study the study whether a potential biomarker can have clinical validity, i.e., whether it can be used for predictive purposes. A successful diagnostic test or biomarker will result in a curve that bends above the diagonal while an unsuccessful test will mirror the diagonal, or fall below it. Thus, a ROC curve provides information whether a diagnostic test is successful or not.

(30) ROC curves in verum patients for activated NKT cells were established (FIG. 1). The area under the curve was determined to be 0.71, which is a clear indication of the reliability of the biomarker. The Youden-index, which can be used to determine an optimum cut-off value for the test's readout, shows an optimum at cut-off at 3.08% activated NKT cells. FIG. 2 shows activated NKT cells versus sensitivity and specificity.

(31) All patients were now sorted into groups depending on their level of activated NKT cells. Patients with cell levels above the cut-off of 3.08% activated NKT cells were designated as biomarker positive, while patients with levels below the cut-off were labelled biomarker negative.

(32) FIG. 3a shows a Kaplan-Meier plot in biomarker positive patients for verum and placebo (solid line: patients treated with MGN1703; dotted line: patients treated with placebo). It is obvious that the survival probability within the biomarker positive group is surprisingly related to the application of the TLR-9 agonist.

(33) FIG. 3b shows a Kaplan-Meier curve for the biomarker negative patients in both the verum and placebo arm. It is obvious that the time of progression free survival is clearly shorter within this verum group as compared to the biomarker positive verum patients (comp. FIG. 3a).

(34) FIG. 4a shows a Kaplan-Meier plot of biomarker positive patients versus biomarker negative patients, only from the verum treatment arm (solid line: biomarker positive patients; dotted line: biomarker negative patients). Clearly, the biomarker positive patients have a significant advantage in survival probability compared to the biomarker negative patients, even when both groups received the verum treatment MGN1703.

(35) FIG. 4b shows a Kaplan-Meier plot of the placebo group for biomarker positive and negative patients (solid line: biomarker positive patients; dotted line: biomarker negative patients). It is obvious that both groups show a more or less identical progression free survival as they were only treated with placebo. Clearly, the biomarker is suitable for assessing whether a patient is a responder to the treatment with a TLR-9 agonist or not. Further, the placebo group shows that the biomarker is not related to effects caused by the overall health status of a patient.

(36) The present invention provides new predictive biomarker for responder to a cancer treatment with a TLR-9 agonist, especially a covalently closed partially self-complementary DNA chain having a double stranded stem an single stranded terminal loops bearing unmethylated CG motifs. The determination of the frequencies of activated NKT cells (CD3+/CD56+/Cd69+) at baseline allows assessing the probability whether a patient is a responder to treatment with the DNA construct or not. Importantly, in the placebo arm, the patients with responder-like characteristics, behave the same way as non-responders did, showing that the prolonged progression-free survival time of the biomarker positive verum patients is in fact due to the applicability of the biomarkers for the selected therapy, not just a better overall health or any non-specific effect.

(37) Material and Methods

(38) Sample Handling

(39) Whole blood (10 mL) for FACS was collected in Streck Cyto-Chex? BCT tubes. Within 2 hours after sampling the blood samples were shipped to the analytical laboratory. According to established Protocol, samples were stored at room temperature. The frequency and activation status of plasmacytoid dendritic cells (pDC), myeloid dendritic cells (mDC), monocytes, natural killer (NK) cells, NKT cells, B cells, T cells and other cell populations were evaluated.

(40) Analytical Methods

(41) Fluorescence activated cell sorting (FACS) were performed according to established principles. Whole blood samples were stained with fluorescence labeled antibodies and incubated. Phenotypical analysis of the immune cells was performed with a FACScalibur (Becton Dickinson) flow cytometer. Frequencies of the respective analyzed cell populations were documented for each sample of the patients.

(42) Analysis of Human PBMC for the Activation of Specific Cell Populations

(43) CD40 Expression of Plasmacytoid Dendritic Cells (pDC)

(44) Cells were stained with the following combination of monoclonal antibodies: Anti-Lineage marker-FITC, (antibody cocktail containing antibodies directed against CD3, CD14, CD16, CD19, CD20, CD56); Anti-CD123-PE; Anti-HLA-DR-PerCP; Anti-CD40-APC; PDC were gated as: lineage negative, HLA-DR positive, CD123 positive cells. Within the PDC population CD40 was used as activation marker.

(45) Activation of NK-, NK-T and T Cells Using the Activation Marker CD69

(46) Cells were stained with the following combination of monoclonal antibodies: Anti CD3-FITC; Anti CD56-PE; Anti CD69-APC

(47) NK cells were gated as: CD3 negative, CD56 positive cells

(48) NK-T cells were gated as: CD3 positive, CD56 positive cells

(49) T cells were gated as: CD3 positive, CD56 negative

(50) CD69 was used as activation marker for all 3 populations.

(51) CD86 Expression of Myeloid Dendritic Cells (MDC):

(52) Cells were stained with the following combination of monoclonal antibodies: Anti-Lineage marker-FITC, (antibody cocktail containing antibodies directed against CD3, CD14, CD16, CD19, CD20, CD56); Anti-CD11c-PE; Anti-HLA-DR-PerCP; Anti-CD86-APC

(53) MDC were gated as: lineage negative, HLA-DR positive, CD11c positive cells.

(54) Within the MDC population CD86 was used as activation marker.

(55) CD86 Expression of B Cells and Monocytes; CD169 Expression of Monocytes

(56) Cells were stained with the following combination of monoclonal antibodies: Anti CD14-FITC; Anti CD169-PE; Anti CD19-PerCP; Anti CD86-APC

(57) B cells were gated as CD19 positive cells. Within the B cell population CD86 was used as activation marker.

(58) Monocytes were gated as CD14 positive cells. Within the monocyte population CD86 and CD169 were used as activation markers.