EP3994280A1

EP3994280A1 - Biomarkers and methods for personalized treatment of small cell lung cancer using kdm1a inhibitors

Info

Publication number: EP3994280A1
Application number: EP19735339.4A
Authority: EP
Inventors: Filippo CICERI; Natalia SACILOTTO; Serena LUNARDI
Original assignee: Oryzon Genomics SA
Current assignee: Oryzon Genomics SA
Priority date: 2019-07-05
Filing date: 2019-07-05
Publication date: 2022-05-11
Also published as: WO2021004610A1; CN114341366A; JP2022546908A

Abstract

The present application discloses biomarkers and methods to predict responsiveness of a patient having small cell lung cancer (SCLC) to treatment with KDM1A inhibitors, and methods of treating subgroups of SCLC patients identified using said methods.

Description

BIOMARKERS AND METHODS FOR PERSONALIZED TREATMENT OF SMALL CELL LUNG CANCER USING

KDM1A INHIBITORS

Field of the invention

The present invention relates to biomarkers and methods for the personalized treatment of small cell lung cancer (SCLC) using KDM1A inhibitors. The invention provides methods to identify patients having SCLC that may benefit from treatment with KDM1 A inhibitors and methods for the treatment of such patients with KDM1A inhibitors.

Background of the invention

Lysine Specific Demethylase 1 (LSD1 , also known as KDM1A) is a histone-modifying enzyme responsible for demethylation of the di-methyl histone 3 lysine 4 (H3K4me2) (Shi et al„ Cell 2004). In several human cancers, KDM1A over-expression has been associated with disease progression and worse prognosis, and its inhibition has been shown to reduce cancer cell growth, migration and invasion. KDM1A has therefore been recognized as a target of interest for the development of new drugs to treat cancer, and several KDM1A inhibitors are currently in clinical trials in oncology.

In particular, KDM1 A inhibitors have been reported to be effective for the treatment of small cell lung cancer (SCLC). It has been shown that KDM1A inhibition reduces proliferation of SCLC cell lines in vitro and delays tumor growth in vivo in SCLC xenograft-bearing mice (Mohammad et al., 2015, Cancer Cell 28, 57-69). However, the published data indicate that while certain SCLC are highly sensitive to KDM1A inhibition, sensitivity to KDM1A inhibition is not a widespread feature of SCLC cells. This raises the need to develop methods of individualizing SCLC treatment with KDM1A inhibitors, and in particular to develop methods of patient selection that allow to identify those patients having SCLC that would benefit, or that would benefit the most, from receiving treatment with a KDM1A inhibitor.

Therefore, the technical problem underlying the present invention is the provision of means and methods for identifying and treating patients having SCLC that are best suited for treatment with a KDM1A inhibitor.

Summary of the invention

The present invention provides means and methods for the personalized treatment of small cell lung cancer (SCLC) using KDM1A inhibitors.

In one aspect, the present invention provides a method of identifying a patient having SCLC who is more likely to respond to a treatment comprising a KDM1A inhibitor, the method comprising measuring the level of ASCL1 and SOX2 in a sample from the patient prior to initiating the treatment comprising a KDM1A inhibitor. In some embodiments, the patient is identified as more likely to respond to a treatment comprising a KDM1A inhibitor when the level of each of ASCL1 and SOX2 in the sample surpasses a threshold.

For example, the patient is identified as more likely to respond to a treatment comprising a KDM1A inhibitor compared to a patient having SCLC and having the level of ASCL1 and SOX2 measured in a sample from the patient prior to initiating the treatment comprising a KDM1A inhibitor when the level of each of ASCL1 and SOX2 in the sample does not surpassa threshold. The latter patient (with the level of each of ASCL1 and SOX2 in the sample not surpassing a threshold) would vice versa be less likely to respond to a treatment comprising a KDM1A inhibitor. This exemplary explanation applies to all aspects and uses provided herein that concern, encompass or comprise identifying a patient having SCLC who is more likely to respond/to be responsive to a treatment comprising a KDM1 A inhibitor or the like.

In some embodiments, the method further comprises using the level of ASCL1 and SOX2 in the sample to generate a score for the sample, wherein the patient is identified as more likely to respond to a treatment comprising a KDM1A inhibitor when the score in the sample surpasses a threshold.

In another aspect, the invention provides a method of identifying a patient having SCLC who may benefit from a treatment comprising a KDM1A inhibitor, the method comprising measuring the level of ASCL1 and SOX2 in a sample from the patient prior to initiating the treatment comprising a KDM1A inhibitor. In some embodiments, the patient is identified as one who may benefit from a treatment comprising a KDM1 A inhibitor when the level of each of ASCL1 and SOX2 in the sample surpasses a threshold. In some embodiments, the method further comprises using the level of ASCL1 and SOX2 in the sample to generate a score for the sample, wherein the patient is identified as one who may benefit from a treatment comprising a KDM1A inhibitor when the score in the sample surpasses a threshold.

In a further aspect, the present invention provides a method of predicting responsiveness of a patient having SCLC to a treatment comprising a KDM1A inhibitor, the method comprising measuring the level of ASCL1 and SOX2 in a sample from the patient prior to initiating the treatment comprising a KDM1A inhibitor. In some embodiments, the patient is identified as more likely to be responsive to a treatment comprising a KDM1A inhibitor when the level of each of ASCL1 and SOX2 in the sample surpasses a threshold. In some embodiments, the method further comprises using the level of ASCL1 and SOX2 in the sample to generate a score for the sample, wherein the patient is identified as more likely to be responsive to a treatment comprising a KDM1A inhibitor when the score in the sample surpasses a threshold.

In a further aspect, the present invention provides a method of assessing the likelihood of a patient having SCLC to respond to a treatment comprising a KDM1A inhibitor, the method comprising measuring the level of ASCL1 and SOX2 in a sample from the patient prior to initiating the treatment comprising a KDM1A inhibitor. In some embodiments, the patient is identified as more likely to respond to a treatment comprising a KDM1A inhibitor when the level of each of ASCL1 and SOX2 in the sample surpasses a threshold. In some embodiments, the method further comprises using the level of ASCL1 and SOX2 in the sample to generate a score for the sample, wherein the patient is identified as more likely to respond to a treatment comprising a KDM1A inhibitor when the score in the sample surpasses a threshold.

In a further aspect, the present invention provides a method of assessing the likelihood of a SCLC in a patient to respond to a treatment comprising a KDM1A inhibitor, the method comprising measuring the level of ASCL1 and SOX2 in a sample from a patient having SCLC prior to initiating the treatment comprising a KDM1A inhibitor. In some embodiments, the SCLC is identified as more likely to respond to a treatment comprising a KDM1 A inhibitor when the level of each of ASCL1 and SOX2 in the sample surpasses a threshold. In some embodiments, the method further comprises using the level of ASCL1 and SOX2 in the sample to generate a score for the sample, wherein the SCLC is identified as more likely to respond to a treatment comprising a KDM1A inhibitor when the score in the sample surpasses a threshold.

In a further aspect, the invention provides a method of selecting a treatment for a patient having SCLC, the method comprising measuring the level of ASCL1 and SOX2 in a sample from the patient prior to initiating the treatment. In some embodiments, the method comprises providing a recommendation that the treatment selected for the patient comprises a KDM1A inhibitor when the level of each of ASCL1 and SOX2 in the sample surpasses a threshold. In some embodiments, the method further comprises using the level of ASCL1 and SOX2 in the sample to generate a score for the sample, and providing a recommendation that the treatment selected for the patient comprises a KDM1A inhibitor when the score in the sample surpasses a threshold. In other words, the treatment to be selected for a patient having SCLC is a treatment comprising a KDM1A inhibitor, for example, when the level of each of ASCL1 and SOX2 in the sample surpasses a threshold or when the score in the sample surpasses a threshold. When the level of each of ASCL1 and SOX2 does not surpass the threshold or when the score does not surpass the threshold, other treatment options than treatment comprising a KDM1A inhibitor may be contemplated for the patients, e.g. treatment with drugs/therapeutic agents other than a KDM1 A inhibitor.

In a further aspect, the invention provides a method of treating a patient having SCLC, the method comprising administering to the patient a therapeutically effective amount of a treatment comprising a KDM1A inhibitor, if the patient has been identified as more likely to respond to a treatment comprising a KDM1A inhibitor using a method according to any of the preceding aspects prior to initiating the treatment comprising a KDM1 A inhibitor.

In a further aspect, the invention features a method of treating a patient having SCLC, the method comprising measuring the level of ASCL1 and SOX2 in a sample from the patient prior to initiating the treatment, identifying the patient as more likely to respond to a treatment comprising a KDM1 A inhibitor when the level of each of ASCL1 and SOX2 in the sample surpasses a threshold, and administering a therapeutically effective amount of a treatment comprising a KDM1A inhibitor to the patient if identified as more likely to respond.

In a further aspect, the invention features a method of treating a patient having SCLC, the method comprising measuring the level of ASCL1 and SOX2 in a sample from the patient prior to initiating the treatment, using these levels to generate a score for the sample, identifying the patient as more likely to respond to a treatment comprising a KDM1A inhibitor when the score in the sample surpasses a threshold, and administering a therapeutically effective amount of a treatment comprising a KDM1 A inhibitor to the patient if identified as more likely to respond.

In a further aspect, the invention provides a KDM1A inhibitor for use in treating a patient having SCLC, wherein the patient has been identified as more likely to respond to a treatment comprising a KDM1A inhibitor using a method according to any of the preceding aspects prior to initiating the treatment comprising a KDM1A inhibitor.

In a further aspect, the invention provides a use of ASCL1 and SOX2 in a method of identifying a patient having SCLC who is more likely to respond to a treatment comprising a KDM1A inhibitor. ln a further aspect, the invention provides a use of ASCL1 and SOX2 in a method of assessing the likelihood of response of a patient having SCLC to a treatment comprising a KDM1A inhibitor.

In a further aspect, the invention provides the use of one or more agents for measuring the level of ASCL1 and SOX2 in a method of identifying a patient having SCLC who is more likely to respond to a treatment comprising a KDM1A inhibitor.

In a further aspect, the invention provides the use of one or more agents for measuring the level of ASCL1 and SOX2 in a method of assessing the likelihood of response of a patient having SCLC to a treatment comprising a KDM1A inhibitor.

In a further aspect, the invention provides a use of ASCL1 and SOX2 for the manufacture of a diagnostic for identifying a patient having SCLC who is more likely to respond to a treatment comprising a KDM1A inhibitor.

In a further aspect, the invention provides a use of ASCL1 and SOX2 for the manufacture of a diagnostic for assessing the likelihood of response of a patient having SCLC to a treatment comprising a KDM1A inhibitor.

In another aspect, the invention provides a kit for identifying a patient having SCLC who is more likely to respond to a treatment comprising a KDM1 A inhibitor, the kit comprising one or more agents for measuring the level of ASCL1 and SOX2 in a sample, and optionally, instructions for use.

In another aspect, the invention provides a kit for assessing the likelihood of response of a patient having SCLC to a treatment comprising a KDM1A inhibitor, the kit comprising one or more agents for measuring the level of ASCL1 and SOX2 in a sample, and optionally, instructions for use.

Brief description of the drawings

Fiqure 1 : Dot plot representing expression of ASCL1 (Y-axis) and SOX2 (X-axis) measured by RNA-seq

(Log2(RPKM) values) for SCLC cell lines sensitive (grey dot), sensitive in part (empty square) or resistant (black triangle) to KDM1A inhibition, as described in more detail in Example 2.

Figure 2: Dot plot representing gene expression measured by qRT-PCR (absolute Cp values) of ASCL1 and

SOX2 for SCLC cell lines sensitive (grey dot), sensitive in part (empty square) or resistant (black diamond) to KDM1A inhibition as described in Example 3. Plotted values are means of independent experiments. The value in brackets has a Cp value above 40 for the expression of SOX2.

Figure 3: Dot plot representing gene expression measured by microarray Affymetrix analysis (RMA values) of

ASCL1 and SOX2 in an extended panel of SCLC cell lines sensitive (grey dot), sensitive in part (empty square), or resistant (black triangle) to KDM1A inhibition, as described in Example 4.

Figure 4: ROC curves based on gene expression of ASCL1 (Figure 4A) and SOX2 (Figure 4B) to discriminate

KDMIAi sensitive and resistant SCLC cells, as described in Example 4. Sensitivity and specificity with their respective confidence interval for a given threshold are indicated in the table below each graph.

Figure 5: Western Blot (WB) (Fig 5A) and quantification (Fig 5B) of ASCL1 and SOX2 protein levels in NCI-H146,

NCI-H510A, NCI-H446 and NCI-H526 cell lines, as described in Example 5.1. Fiqure 6: Correlation between protein and mRNA (CCLE Affymetrix) levels for SOX2 (Fig 6A) and ASCL1 (Fig

6B), as described in Example 5.1.

Figure 7: Fluorescent immunohistochemistry stainings for SOX2 (Fig. 7A), ASCL1 (Fig. 7B) and negative control

(no primary antibody, Fig. 7C) according to Example 5.2, showing low/undetectable levels of both biomarkers in NCI-H526 cells, low/undetectable levels of ASCL1 in NCI-H446 cells and the highest levels for both biomarkers in NCI-H146 cells, in agreement with their mRNA levels. DAPI: DAPI (4',6- diamidino-2-phenylindole) is a fluorescent dye that strongly binds to A-T rich regions in DNA and stains the nuclei. Signal is not detected in negative control with secondary antibody only (AF546: Alexa Fluor 546).

Figure 8: Correlation between protein and mRNA (CCLE Affymetrix) levels for SOX2 (Fig 8A) and ASCL1 (Fig

8B), as described in more detail in Example 5.2.

Figure 9: Representative ASCL1 , SOX2 and their corresponding DAPI fluorescent immunohistochemistry images from SCLC PDX TMA are shown for each staining classification level 0, 1 , 2 and 3, as described in Example 5.3

Figure 10: Two-tailed Spearman correlation tests between RNASeq (Log2FPKM) and IF (visual score) datasets for SOX2 (Figure 10A and 10B) and ASCL1 (Figure 10C and 10D) with a 95% confidence interval are shown for two independent experiments (coefficients are specified at the bottom of each graph), as described in Example 5.3.

Figure 11 : WB and Ponceau staining for ASCL1 and SOX2 in exosomal fraction according to Example 6. In agreement with mRNA expression, ASCL1 was not detected in exosomes derived from NCI-H446 and NCI-H526 and SOX2 was not present in exosomes derived from NCI-H526.

Figure 12: WB (left) and Ponceau staining (right) for ASCL1 , SOX2 and CD151 (lung cancer-specific exosome marker) in both exosomes and the corresponding parental cells, according to Example 6. ASCL1 , SOX2 and CD151 signals are significantly reduced or ablated in the exosomal fraction after treatment of NCI-H510A cells with 5mM GW4869 (exosome production inhibitor) for 48 hours, while expression of these proteins in cells is not affected, indicating detection of ASCL1 ,SOX2 and CD151 is exosome- specific.

Detailed description of the invention

Definitions

Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Although methods and materials similar or equivalent to those described herein can be used in the practice or testing of the invention, suitable methods and materials are described below. All publications, patent applications, patents, and other references mentioned herein are incorporated by reference in their entirety.

The nomenclature used in this Application is based on lUPAC systematic nomenclature, unless indicated otherwise.

Any open valency appearing on a carbon, oxygen, sulfur or nitrogen atom in the structures herein indicates the presence of a hydrogen, unless indicated otherwise.

Stereochemical definitions and conventions used herein generally follow S. P. Parker, Ed., McGraw-Hill Dictionary of Chemical Terms (1984) McGraw-Hill Book Company, New York; and Eliel, E. and Wilen, S., “Stereochemistry of Organic Compounds”, John Wiley & Sons, Inc., New York, 1994. In describing an optically active compound, the prefixes D and L, or R and S, are used to denote the absolute configuration of the molecule about its chiral center(s). The substituents attached to the chiral center under consideration are ranked in accordance with the Sequence Rule of Cahn, Ingold and Prelog, (Cahn et al. Angew. Chem. Inter. Edit. 1966, 5, 385; errata 511). The prefixes D and L or (+) and (-) are employed to designate the sign of rotation of plane-polarized light by the compound, with (-) or L designating that the compound is levorotatory. A compound prefixed with (+) or D is dextrorotatory.

The term“optional” or“optionally” denotes that a subsequently described event or circumstance can but need not occur, and that the description includes instances where the event or circumstance occurs and instances in which it does not.

The term“pharmaceutically acceptable salts” denotes salts which are not biologically or otherwise undesirable. Pharmaceutically acceptable salts include both acid and base addition salts. Pharmaceutically acceptable salts are well known in the art.

The term“pharmaceutically acceptable acid addition salt” denotes those pharmaceutically acceptable salts formed with inorganic acids such as hydrochloric acid, hydrobromic acid, sulfuric acid, nitric acid, carbonic acid, phosphoric acid, and organic acids selected from aliphatic, cycloaliphatic, aromatic, araliphatic, heterocyclic, carboxylic, and sulfonic classes of organic acids such as formic acid, acetic acid, propionic acid, glycolic acid, gluconic acid, lactic acid, pyruvic acid, oxalic acid, malic acid, maleic acid, maloneic acid, succinic acid, fumaric acid, tartaric acid, citric acid, aspartic acid, ascorbic acid, glutamic acid, anthranilic acid, benzoic acid, cinnamic acid, mandelic acid, embonic acid, phenylacetic acid, methanesulfonic acid, ethanesulfonic acid, p-toluenesulfonic acid, and salicyclic acid.

The term“pharmaceutically acceptable base addition salt” denotes those pharmaceutically acceptable salts formed with an organic or inorganic base. Examples of acceptable inorganic bases include sodium, potassium, ammonium, calcium, magnesium, iron, zinc, copper, manganese, and aluminum salts. Salts derived from pharmaceutically acceptable organic nontoxic bases includes salts of primary, secondary, and tertiary amines, substituted amines including naturally occurring substituted amines, cyclic amines and basic ion exchange resins, such as isopropylamine, trimethylamine, diethylamine, triethylamine, tripropylamine, ethanolamine, 2- diethylaminoethanol, trimethamine, dicyclohexylamine, lysine, arginine, histidine, caffeine, procaine, hydrabamine, choline, betaine, ethylenediamine, glucosamine, methylglucamine, theobromine, purines, piperizine, piperidine, N- ethylpiperidine, and polyamine resins.

The term“KDM1A inhibitor” or“KDM1 Ai” are used interchangeably and as used herein means any compound that is capable of inhibiting KDM1A activity. Methods to determine KDM1A inhibitory activity are well known in the art. In preferred embodiments, the KDM1A inhibitor is a small molecule. Examples of KDM1A inhibitors are described in more detail elsewhere herein.

As used herein, a“small molecule" refers to an organic compound with a molecular weight below 900 daltons, preferably below 500 daltons. The molecular weight is the mass of a molecule and is calculated as the sum of the atomic weights of each constituent element multiplied by the number of atoms of that element in the molecular formula.

A“treatment comprising a KDM1A inhibitor” means any therapy or treatment regimen incorporating a KDM1A inhibitor, whether as a sole active pharmaceutical ingredient (API) or in combination with one or more additional APIs, like other anticancer agents. Said treatment comprising a KDM1A inhibitor will typically be in the form of a pharmaceutical composition. In case the treatment comprising a KDM1A inhibitor comprises one or more APIs in addition to the KDM1A inhibitor, they may be administered in the form of a single pharmaceutical composition incorporating all APIs, or else may be administered in the form of individual pharmaceutical compositions for each API (i.e. the KDM1A inhibitor and the one or more additional APIs), which may be administered by the same or different routes (e.g. one may be administered orally and the other one parenterally), and which may be administered simultaneously or sequentially.

The terms“pharmaceutical composition" and “pharmaceutical formulation" are used interchangeably and denote a composition (e.g. a mixture or solution) comprising a therapeutically effective amount of an active pharmaceutical ingredient (e.g. a KDM1A inhibitor) together with one or more pharmaceutically acceptable excipients to be administered to a mammal, e.g. a human in need thereof.

The term“pharmaceutically acceptable” denotes an attribute of a material which is useful in preparing a pharmaceutical composition that is generally safe, non-toxic, and neither biologically nor otherwise undesirable and is acceptable for veterinary as well as human pharmaceutical use.

The terms“pharmaceutically acceptable excipient”,“pharmaceutically acceptable carrier” and“therapeutically inert excipient” can be used interchangeably and denote any pharmaceutically acceptable ingredient in a pharmaceutical composition having no therapeutic activity and being non-toxic to the subject administered, such as disintegrators, binders, fillers, solvents, buffers, tonicity agents, stabilizers, antioxidants, surfactants, carriers, diluents or lubricants used in formulating pharmaceutical products.

The term“therapeutically effective amount” (or“effective amount”) denotes an amount of a compound of the present invention that, when administered to a patient, (i) treats or prevents the particular disease, (ii) attenuates, ameliorates or eliminates one or more symptoms of the disease, or (iii) prevents or delays the onset of one or more symptoms of the disease. The therapeutically effective amount will vary depending on the compound, the disease state being treated, the severity of the disease treated, the age and relative health of the patient, the route and form of administration, the judgment of the attending medical or veterinary practitioner, and other factors.

The term“treating" or“treatment" of a disease (e.g. SCLC) includes reversing, alleviating, or inhibiting the progress of the disease or one or more symptoms thereof.

A“patient” or“subject" may be used interchangeably, and means a mammal in need of treatment. Mammals include, but are not limited to primates (e.g., humans and non-human primates such as monkeys), domesticated animals (e.g., cows, sheep, cats, dogs, and horses) and laboratory animals (mice, rats, guinea pigs and the like). In preferred embodiments, the patient is a human. Intended to be included as a patient are any subjects involved in clinical research trials. The patient may have been previously treated for example with other drugs and/or with any KDM1A inhibitor. In one aspect, the patient has not been previously treated with any KDM1A inhibitor. The patient may be being treated with other drugs, particularly at the time of obtaining the sample, but shall not be being treated with any KDM1A inhibitor at the time of obtaining the sample for use in the methods according to the invention (i.e. the patient should not be treated concurrently with a KDM1A inhibitor at the time of obtaining the sample). Alternatively, the patient shall not be being treated with any KDM1A inhibitor within a period of time prior to obtaining the sample, if biomarker levels could be still modulated by the (remaining) KDM1A inhibitor within said period of time (for example if the biomarker levels have not returned to the level before a previous treatment with (or administration of) a KDM1A inhibitor). For example, the patient shall not be being treated with any KDM1A inhibitor within two weeks, or more preferably within one month, prior to obtaining the sample. The latter is to avoid that biomarker levels could be still modulated by the (remaining) KDM1A inhibitor.

The term“biomarker” or“marker” as used herein refers to a protein or polynucleotide, the expression or presence of which in or on a mammalian tissue or cell can be detected by standard methods (or methods disclosed herein) and which is associated with a mammalian cell’s or tissue’s sensitivity to treatment comprising a KDM1A inhibitor. The biomarkers according to the invention are ASCL1 and SOX2,

The term“measuring” the level of the biomarker as used herein refers to experimentally determining the amount of biomarker in the sample, employing appropriate methods of detection as described elsewhere herein.

The term“threshold” as used herein refers to a predetermined value, line or a more complex n dimensional function that defines the frontier between two categories/subsets of a population, e.g. patients with SCLC more likely to respond to KDM1A inhibitor treatment vs. patients with SCLC less likely to respond to KDM1A inhibitor treatment. In the methods according to the invention, different thresholds can apply to individual biomarker levels (i.e. each biomarker, ASCL1 and SOX2, has its respective threshold) or to a score derived from the levels of the biomarkers by application of a classification algorithm as described elsewhere herein. As the skilled person will appreciate, thresholds are established to optimally distinguish between samples of different categories. Thresholds can be established according to methods known in the art. Typically, a threshold can be determined experimentally or theoretically using a training set of samples with known sensitivity or resistance to treatment with KDM1A inhibitors. Training samples can be e.g. SCLC cell lines, patient-derived xenografts (PDX) or human clinical samples with known sensitivity or resistance to treatment with a KDM1A inhibitor. A threshold can also be arbitrarily selected based upon existing experimental and/or clinical and/or regulatory requirements, as would be recognized by a person of ordinary skill in the art. Preferentially the threshold is established in order to obtain the optimal sensitivity and specificity according to the function of the test and the benefit/risk balance (clinical consequences of false positive and false negative). Typically, the optimal sensitivity and specificity (and so the threshold) can be determined using a Receiver Operating Characteristic (ROC) curve based on experimental data, as shown in the appended Examples. In some embodiments, the threshold is a threshold value. In some embodiments, threshold values for the biomarkers are derived from the ASCL1 and SOX2 levels measured in one or more samples of patient-derived SCLC cells that are sensitive or resistant to treatment comprising a KDM1A inhibitor. In some embodiments, threshold values are derived from the ASCL1 and SOX2 levels measured in one or more samples of (human) patients that have responded or not responded to treatment comprising a KDM1A inhibitor. In some embodiments, the threshold values are derived from the ASCL1 and SOX2 levels measured in one or more samples obtained from patient derived xenograft models that have responded or not responded to treatment comprising a KDM1A inhibitor. In some embodiments, the threshold values are obtained from the mRNA levels of the biomarkers. In some embodiments, the threshold values are obtained from the protein levels of the biomarkers.

The term“score’’ as used herein refers to the output calculated from the biomarker levels measured in a sample by a classification algorithm. The score will be/is compared to a threshold and used to decide whether a patient from which a sample is derived is more likely or less likely to respond to a treatment comprising a KDM1A inhibitor. For example, when the score surpasses a threshold, the patient is identified as more likely to respond to a treatment comprising a KDM1A inhibitor / the patient is identified that he/she may benefit from treatment comprising a KDM1A inhibitor.

A“classification algorithm” as used herein is a mathematical function that is used to calculate a score for a sample and evaluate (“classify”) which category the sample belongs to, i.e. if it surpasses a threshold. Classification algorithms are well known in the art. Examples of classification algorithms include: linear classifiers, fisher's linear discriminant, linear Boolean classification, logistic regression, naive bayes classifier, perceptron, support vector machines, least squares support vector machines, quadratic classifiers, kernel estimation, k-nearest neighbor, decision trees, random forests, neural networks, learning vector quantization. Software packages incorporating one or several classification algorithms are readily available for use online or for download and include, for example, DTREG, XLStat, http://www.support-vector-machines.org/SVM_soft.html, and the like. In some embodiments, the classification algorithm is a Boolean function (truth function).

Samples can be classified using the Boolean conjunction function A AND B, wherein A and B evaluate whether the level of each of the biomarkers (ASCL1 and S0X2) in the sample are above that biomarker’s respective threshold. The Boolean conjunction function yields a score, typically represented by 1 (truth) when all criteria are complied with (for example if the level of each of the biomarkers (ASCL1 and S0X2) in the sample is above/surpasses that biomarker’s respective threshold(s)), or 0 (falsehood) when one (or both) of the criteria is not (for example if only one level or none level of each of the biomarkers (ASCL1 and S0X2) in the sample is above/surpasses that biomarker’s respective threshold(s)). The threshold applied to the score generated by the Boolean algorithm to classify the sample is 0, i.e. samples surpassing this threshold (i.e. with score > 0 ) are classified as likely to respond to treatment comprising a KDM1A inhibitor / to benefit from treatment comprising a KDM1A inhibitor.

In some embodiments, the classification algorithm is a Support Vector Machine (SVM). SVM is used to perform a classification by mapping a training data set in space and constructing an N-dimensional hyperplane that optimally separates the sample data into two categories (e.g. sensitive and resistant to KDMIAi) and thus acts as a threshold function. In addition to performing linear classification, SVMs can perform a non-linear classification using the kernel trick, mapping the sample data into high-dimensional feature spaces. Using the scoring function of the trained SVM, new data are then mapped into that same space and predicted to belong to a category based on which side of the hyperplane they fall. The performance of a classification algorithm (threshold included) can be further evaluated by comparing the classification predicted by the algorithm with experimental values and calculating true positives, false positives, true negatives, and false negatives as well as the sensitivity, specificity, etc and may optionally be subjected to multiple rounds of training using training samples of known responsiveness/resistance to KDMIAi in order to tune model parameters and/or optimise performance.

The term“surpass” as used herein means: A biomarker level or score of a test sample (with unknown sensitivity to KDMIAi), e.g. a sample from a SCLC patient that is being considered for receiving a treatment comprising a KDMIAi, will surpass or cross a threshold when the comparison of the levels or score (as the case may be) of that test sample with the respective threshold classifies that sample in the category of samples known to be sensitive to KDMIAi. In some embodiments, the threshold is a threshold value, and the biomarker level or score will surpass the threshold (threshold value) when the biomarker level or score is above its respective threshold value. The comparison of the level of the biomarker or score in the sample with the respective threshold for the biomarker or score may be carried out mentally, manually or can be automatically carried out by a computer program.

The term“sample” as used herein in relation to a patient sample to be used in the methods according to the invention can be a tumor sample (e.g. a biopsy sample, such as a biopsy sample either from primary or metastatic SCLC lesions), a body fluid or a patient derived cell line, a PDX sample (“PDX” means a“patient-derived xenograft", i.e. a human tumor grown in mice) or (an)exosome(s). Preferably, the sample is rich in/enriched for tumor cells. The sample from the patient to be used to practice the methods according to the present invention is to be obtained prior to initiating treatment with a KDM1A inhibitor (i.e. in absence of current treatment with a KDM1A inhibitor and/or not during, for example, a period of time following previous treatment with (administration of) a KDM1A inhibitor, if biomarker levels could be still modulated by the (remaining) KDM1A inhibitor within that period - For example, the sample from the patient to be used in accordance with the methods according to the present invention is to not be obtained within two weeks, preferably one month, following previous treatment with (or administration of) a KDM1A inhibitor). Biopsy samples can be obtained by well-known techniques and may be fresh or may be subjected to post- collection preparative and storage techniques (e.g. frozen, fixed and/or embedded, such as for example formalin- fixed, paraffin-embedded, fresh snap-frozen, fixed and frozen OCT-embedded, etc). Samples of body fluids can be obtained by well-known techniques and include samples of blood, sputum, bronchoalveolar lavage fluid or any other bodily secretion or derivative thereof that may contain SCLC cells. Separated cells may be obtained from the body fluids or the tissues or organs by separating techniques such as centrifugation or cell sorting. The cell sample can, of course, be subjected to a variety of well-known post-collection preparative and storage techniques (e.g., nucleic acid and/or protein extraction, fixation, storage, freezing, ultrafiltration, concentration, evaporation, centrifugation, etc.) prior to assessing the level of the markers in the sample. Preferably the sample in which the biomarker levels are measured is rich in / enriched for the presence of SCLC cells or for SCLC cell-derived vesicles (e.g. exosomes, etc). For example, SCLC cells can be isolated from sputum using methods described in the literature (Chest. 1992 Aug;102(2);372-4). SCLC circulating tumor cells (CTCs) can be purified from blood by methods described in the literature (Peeters et al., Br J Cancer 2013 Apr 2; 108(6) : 1358-67 ; Hodgkinson et al., Nat Med. 2014 Aug;20(8):897- 903; Carter et al, Nat Med, 2017 Jan;23(1):114-119) and SCLC derived exosomes can be purified from blood by various methods described in the literature (Li et al.Jheranostics. 2017; 7(3): 789-804; Sandfeld-Paulsen et al.,J Thorac Oncol. 2016 Oct; 11(10): 1701 -10). The biomarker levels can also be analyzed in a spatially defined area from the sample rich in SCLC cells, as may be determined e.g. by anatomopathologists using standard methods used in the field.

The terms “responsiveness to”, “responsive to”, “respond to”, “sensitivity to”, “sensitive to” and the like expressions in the context of a treatment comprising a KDM1A inhibitor mean that a patient having SCLC (or a sample, SCLC cell line, etc) shows a positive response to KDM1A inhibition, i.e. to a treatment comprising a KDM1A inhibitor. In a more simplified form, the terms“responsive to a treatment comprising a KDM1A inhibitor” and the like can be phrased as“responsive to a KDM1A inhibitor”,“responsive to KDM1A inhibition” and the like. For example,“a positive response to a treatment comprising a KDM1A inhibitor" or“a benefit from a treatment comprising a KDM1A inhibitor” can be or can include reversing, alleviating, or inhibiting the progress of the disease SCLC or one or more symptoms thereof. The terms“more likely to respond” as used herein can mean“more responsive to” or, simply, “responsive to”.

The phrase“identifying a patient” or“selecting a patient” may be used interchangeably and as used herein refers to using the information or data generated relating to the biomaker levels in a sample of a patient to identify or selecting the patient as more likely to respond to (or to benefit from) or less likely to respond to (or to benefit from) a treatment comprising a KDM1A inhibitor. The information or data used or generated may be in any form, written, oral or electronic. The herein provided methods and uses can include communicating the result, information or data to the patient and/or any person(s) involved in or in charge of the treatment of the patient, the treatment comprising a KDM1 A inhibitor. In some embodiments, using the information or data generated includes communicating, presenting, reporting, storing, sending, transferring, supplying, transmitting, dispensing, or combinations thereof. In some embodiments, communicating, presenting, reporting, storing, sending, transferring, supplying, transmitting, dispensing, or combinations thereof are performed by a computing device, analyzer unit or combination thereof. In some further embodiments, communicating, presenting, reporting, storing, sending, transferring, supplying, transmitting, dispensing, or combinations thereof are performed by a laboratory or medical professional. In some embodiments, the information or data includes a comparison of the biomarker levels to a threshold. In some embodiments, the information or data includes an indication that the patient is more likely or less likely to respond to (or to benefit from) a treatment comprising a KDM1 A inhibitor.

The phrase“predicting responsiveness of a patient” as used herein refers to using the information or data generated relating to the biomaker levels in a sample of a patient to evaluate the likelihood that the patient will respond to a treatment comprising a KDM1 A inhibitor. The information or data used or generated may be in any form, written, oral or electronic. In some embodiments, using the information or data generated includes communicating, presenting, reporting, storing, sending, transferring, supplying, transmitting, dispensing, or combinations thereof. In some embodiments, communicating, presenting, reporting, storing, sending, transferring, supplying, transmitting, dispensing, or combinations thereof are performed by a computing device, analyzer unit or combination thereof. In some further embodiments, communicating, presenting, reporting, storing, sending, transferring, supplying, transmitting, dispensing, or combinations thereof are performed by a laboratory or medical professional. In some embodiments, the information or data includes a comparison of the biomarker levels to a threshold. In some embodiments, the information or data includes an indication that the patient is more likely or less likely to respond to a treatment comprising a KDM1A inhibitor.

The phrase“selecting a treatment’’ as used herein refers to using the information or data generated relating to the biomarker levels in a sample of a patient to identify or selecting a treatment (therapy) for a patient. In some embodiment the treatment may comprise a KDM1A inhibitor. The information or data used or generated may be in any form, written, oral or electronic. In some embodiments, using the information or data generated includes communicating, presenting, reporting, storing, sending, transferring, supplying, transmitting, dispensing, or combinations thereof. In some embodiments, communicating, presenting, reporting, storing, sending, transferring, supplying, transmitting, dispensing, or combinations thereof are performed by a computing device, analyzer unit or combination thereof. In some further embodiments, communicating, presenting, reporting, storing, sending, transferring, supplying, transmitting, dispensing, or combinations thereof are performed by a laboratory or medical professional. In some embodiments, the information or data includes a comparison of biomarker levels to a threshold. In some embodiments, the information or data includes an indication that a treatment comprising a KDM1A inhibitor is suitable for the patient (i.e. the patient is likely to respond to said treatment).

The phrase "recommending a treatment” as used herein refers to using the information or data generated relating to the biomarker levels for proposing or selecting a treatment comprising a KDM1A inhibitor for a patient identified or selected as more or less likely to respond to the treatment comprising a KDM1A inhibitor. The information or data used or generated may be in any form, written, oral or electronic. In some embodiments, using the information or data generated includes communicating, presenting, reporting, storing, sending, transferring, supplying, transmitting, dispensing, or combinations thereof. In some embodiments, communicating, presenting, reporting, storing, sending, transferring, supplying, transmitting, dispensing, or combinations thereof are performed by a computing device, analyzer unit or combination thereof. In some further embodiments, communicating, presenting, reporting, storing, sending, transferring, supplying, transmitting, dispensing, or combinations thereof are performed by a laboratory or medical professional. In some embodiments, the information or data includes a comparison of the biomarker levels to a threshold. In some embodiments, the information or data includes an indication that a treatment comprising a KDM1 A inhibitor is suitable for the patient.

As used herein, a“kit" is any manufacture (e.g. a package or container) comprising one or more agents for measuring the level of ASCL1 and SOX2, as described herein, the manufacture being promoted, distributed or sold as a unit for performing the methods of the present invention.

As used herein,“agents” for measuring the level of ASCL1 or SOX2 include any reagents typically used in the field for measuring biomarker levels, including without limitation antibodies that specifically recognize ASCL1 or SOX2 proteins, or probes and/or primers that hybridize with ASCL1 or SOX2 polynucleotides, for specifically detecting the biomarkers according to the invention, and any other such reagent as described in more detail elsewhere herein in connection with the methods for measuring biomarker levels.

The present invention provides means and methods for identifying patients having SCLC with increased likelihood to respond to treatment with KDM1A inhibitors and thus that are best suited for treatment comprising a KDM1 A inhibitor, and therapeutic methods for treating those patients with KDM1A inhibitors. The invention is based, at least in part, on the discovery that levels of ASCL1 and SOX2 can be used as biomarkers (e.g. predictive biomarkers) in methods of predicting likelihood to respond to treatment with KDM1A inhibitors. As documented herein and in the appended Examples, the inventors have found that high levels of both ASCL1 and SOX2 in SCLC cell lines correlate with responsiveness (sensitivity) of SCLC to KDM1A inhibitor therapy. As shown in Examples 2, 3 and 4 and in Figures 1 to 3, SCLC cell lines that express high levels of both ASCL1 and SOX2 are usually responsive (sensitive) to KDM1A inhibitors, whereas SCLC cell lines that exhibit low levels of either one or both of ASCL1 and SOX2 are usually resistant to KD 1A inhibition treatment. ASCL1 and SOX2 levels can thus be used to stratify SCLC patients for treatment with KDM1A inhibitors, identifying those patients that are more likely to be responsive to treatment with a KDM1A inhibitor from those that are less likely to respond to KDM1A inhibition. The methods according to the invention using ASLC1 and SOX2 are able to predict responsiveness of SCLC to KDM1A inhibition with high sensitivity and specificity, as shown in more detail in the appended Examples, which is remarkable in view of the reduced number of biomarkers used. The methods according to the invention are amenable to measuring the biomarkers as either mRNA levels or protein levels, since a good correlation has been shown between their mRNA and protein expression levels, as shown in Example 5 using either SCLC cell lines or patient-derived samples like SCLC PDX samples. This makes the methods according to the invention particularly advantageous for use in clinical practice, in particular for use in hospitals, as samples most readily available from cancer patients are typically fixed tumor biopsies, which are better suited for protein level analysis because standard biopsy sample fixing procedures are known to fragment and reduce RNA.

As indicated previously, the methods of the invention comprise measuring the level of ASCL1 and SOX2. These markers per se are known in the art and also described herein below.

Public data base entries for ASCL1 and SOX2: DNA and protein sequences of human ASCL1 and SOX2 have been previously reported, see GenBank Numbers (NCBI-GenBank Flat File Release 225.0, April 15, 2018) and UniProtKB/Swiss-Prot Numbers (Knowledgebase Release 2018J04, April 25, 2018) listed below, each of which is incorporated herein by reference in its entirety for all purposes. Such sequences can be used to design procedures for measuring the level of ASCL1 and SOX2 by ways known to one skilled in the art.

Exemplary nucleotide sequences and amino acid sequences of human ASCL1 and SOX2 are shown in SEQ ID NO: 1 to 4 herein. The following table allocates the markers and the respective sequences:

In one aspect, the present invention provides a method of identifying a patient having SCLC who is more likely to respond to a treatment comprising a KDM1A inhibitor, the method comprising measuring the level of ASCL1 and SOX2 in a sample from the patient prior to initiating the treatment comprising a KDM1A inhibitor. In some embodiments, the patient is identified as more likely to respond to a treatment comprising a KDM1A inhibitor when the level of each of ASCL1 and SOX2 in the sample surpasses a threshold. In some embodiments, the method further comprises using the level of ASCL1 and SOX2 in the sample to generate a score for the sample, wherein the patient is identified as more likely to respond to a treatment comprising a KDM1A inhibitor when the score in the sample surpasses a threshold. Accordingly, in some embodiments, the invention provides a method of identifying a patient having SCLC who is more likely to respond to a treatment comprising a KDM1A inhibitor, the method comprising measuring the level of ASCL1 and SOX2 in a sample from the patient prior to initiating the treatment comprising a KDM1A inhibitor, wherein the patient is identified as more likely to respond to a treatment comprising a KD 1A inhibitor when the level of each of ASCL1 and SOX2 in the sample surpasses a threshold. In some embodiments, the invention provides a method of identifying a patient having SCLC who is more likely to respond to a treatment comprising a KDM1A inhibitor, the method comprising measuring the level of ASCL1 and SOX2 in a sample from the patient prior to initiating the treatment comprising a KDM1A inhibitor, and using the level of ASCL1 and SOX2 in the sample to generate a score for the sample, wherein the patient is identified as more likely to respond to a treatment comprising a KDM1 A inhibitor when the score in the sample surpasses a threshold.

In another aspect, the invention provides a method of identifying a patient having SCLC who may benefit from a treatment comprising a KDM1A inhibitor, the method comprising measuring the level of ASCL1 and SOX2 in a sample from the patient prior to initiating the treatment comprising a KDM1A inhibitor. In some embodiments, the patient is identified as one who may benefit from a treatment comprising a KDM1A inhibitor when the level of each of ASCL1 and SOX2 in the sample surpasses a threshold. In some embodiments, the method further comprises using the level of ASCL1 and SOX2 in the sample to generate a score for the sample, wherein the patient is identified as one who may benefit from a treatment comprising a KDM1A inhibitor when the score in the sample surpasses a threshold. Accordingly, in some embodiments, the invention features a method of identifying a patient having SCLC who may benefit from a treatment comprising a KDM1A inhibitor, the method comprising measuring the level of ASCL1 and SOX2 in a sample from the patient prior to initiating the treatment comprising a KDM1A inhibitor, wherein the patient is identified as one who may benefit from a treatment comprising a KDM1A inhibitor when the level of each of ASCL1 and SOX2 in the sample surpasses a threshold. In some embodiments, the invention features a method of identifying a patient having SCLC who may benefit from a treatment comprising a KDM1A inhibitor, the method comprising measuring the level of ASCL1 and SOX2 in a sample from the patient prior to initiating the treatment comprising a KDM1A inhibitor, and using the level of ASCL1 and SOX2 in the sample to generate a score for the sample, wherein the patient is identified as one who may benefit from a treatment comprising a KDM1A inhibitor when the score in the sample surpasses a threshold.

In a further aspect, the present invention provides a method of predicting responsiveness of a patient having SCLC to a treatment comprising a KDM1A inhibitor, the method comprising measuring the level of ASCL1 and SOX2 in a sample from the patient prior to initiating the treatment comprising a KDM1A inhibitor. In some embodiments, the patient is identified as more likely to be responsive to a treatment comprising a KDM1A inhibitor when the level of each of ASCL1 and SOX2 in the sample surpasses a threshold. In some embodiments, the method further comprises using the level of ASCL1 and SOX2 in the sample to generate a score for the sample, wherein the patient is identified as more likely to be responsive to a treatment comprising a KDM1A inhibitor when the score in the sample surpasses a threshold. Accordingly, in some embodiments, the present invention provides a method of predicting responsiveness of a patient having SCLC to a treatment comprising a KDM1A inhibitor, the method comprising measuring the level of ASCL1 and SOX2 in a sample from the patient prior to initiating the treatment comprising a KDM1A inhibitor, wherein the patient is identified as more likely to be responsive to a treatment comprising a KDM1A inhibitor when the level of each of ASCL1 and SOX2 in the sample surpasses a threshold. In some embodiments, the present invention provides a method of predicting responsiveness of a patient having SCLC to a treatment comprising a KDM1 A inhibitor, the method comprising measuring the level of ASCL1 and SOX2 in a sample from the patient prior to initiating the treatment comprising a KDM1A inhibitor, and using the level of ASCL1 and SOX2 in the sample to generate a score for the sample, wherein the patient is identified as more likely to be responsive to a treatment comprising a KDM1 A inhibitor when the score in the sample surpasses a threshold.

In a further aspect, the present invention provides a method of assessing the likelihood of a patient having SCLC to respond to a treatment comprising a KDM1A inhibitor, the method comprising measuring the level of ASCL1 and SOX2 in a sample from the patient prior to initiating the treatment comprising a KDM1A inhibitor. In some embodiments, the patient is identified as more likely to respond to a treatment comprising a KDM1A inhibitor when the level of each of ASCL1 and SOX2 in the sample surpasses a threshold. In some embodiments, the method further comprises using the level of ASCL1 and SOX2 in the sample to generate a score for the sample, wherein the patient is identified as more likely to respond to a treatment comprising a KDM1A inhibitor when the score in the sample surpasses a threshold. Accordingly, in some embodiments, the present invention provides a method of assessing the likelihood of a patient having SCLC to respond to a treatment comprising a KDM1A inhibitor, the method comprising measuring the level of ASCL1 and SOX2 in a sample from the patient prior to initiating the treatment comprising a KDM1A inhibitor, wherein the patient is identified as more likely to respond to a treatment comprising a KDM1A inhibitor when the level of each of ASCL1 and SOX2 in the sample surpasses a threshold. In some embodiments, the present invention provides a method of assessing the likelihood of a patient having SCLC to respond to a treatment comprising a KDM1A inhibitor, the method comprising measuring the level of ASCL1 and SOX2 in a sample from the patient prior to initiating the treatment comprising a KDM1A inhibitor, and using the level of ASCL1 and SOX2 in the sample to generate a score for the sample, wherein the patient is identified as more likely to respond to a treatment comprising a KDM1A inhibitor when the score in the sample surpasses a threshold.

In a further aspect, the present invention provides a method of assessing the likelihood of a SCLC in a patient to respond to a treatment comprising a KDM1A inhibitor, the method comprising measuring the level of ASCL1 and SOX2 in a sample from a patient having SCLC prior to initiating the treatment comprising a KDM1 A inhibitor. In some embodiments, the SCLC is identified as more likely to respond to a treatment comprising a KDM1A inhibitor when the level of each of ASCL1 and SOX2 in the sample surpasses a threshold. In some embodiments, the method further comprises using the level of ASCL1 and SOX2 in the sample to generate a score for the sample, wherein the SCLC is identified as more likely to respond to a treatment comprising a KDM1A inhibitor when the score in the sample surpasses a threshold. Accordingly, in some embodiments, the present invention provides a method of assessing the likelihood of a SCLC to respond to a treatment comprising a KDM1A inhibitor, the method comprising measuring the level of ASCL1 and SOX2 in a sample from a patient having SCLC prior to initiating the treatment comprising a KDM1A inhibitor, wherein the SCLC is identified as more likely to respond to a treatment comprising a KDM1A inhibitor when the level of each of ASCL1 and SOX2 in the sample surpasses a threshold. In some embodiments, the present invention provides a method of assessing the likelihood of a SCLC to respond to a treatment comprising a KDM1A inhibitor, the method comprising measuring the levels of ASCL1 and SOX2 in a sample from a patient having SCLC prior to initiating the treatment comprising a KDM1A inhibitor, and using the level of ASCL1 and SOX2 in the sample to generate a score for the sample, wherein the SCLC is identified as more likely to respond to a treatment comprising a KDM1 A inhibitor when the score in the sample surpasses a threshold.

In a further aspect, the invention provides a method of selecting a treatment for a patient having SCLC, the method comprising measuring the level of ASCL1 and SOX2 in a sample from the patient prior to initiating the treatment. In some embodiments, the method comprises providing a recommendation that the treatment selected for the patient comprises a KDM1A inhibitor when the level of each of ASCL1 and SOX2 in the sample surpasses a threshold. In some embodiments, the method further comprises using the level of ASCL1 and SOX2 in the sample to generate a score for the sample, and providing a recommendation that the treatment selected for the patient comprises a KDM1A inhibitor when the score in the sample surpasses a threshold. Accordingly, in some embodiments, the invention further provides a method of selecting a treatment for a patient having SCLC, the method comprising measuring the level of ASCL1 and SOX2 in a sample from the patient prior to initiating the treatment, and providing a recommendation that the treatment selected for the patient comprises a KDM1A inhibitor when the level of each of ASCL1 and SOX2 in the sample surpasses a threshold. In some embodiments, the invention further provides a method of selecting a treatment for a patient having SCLC, the method comprising measuring the level of ASCL1 and SOX2 in a sample from the patient prior to initiating the treatment, using the level of ASCL1 and SOX2 in the sample to generate a score for the sample, and providing a recommendation that the treatment selected for the patient comprises a KDM1A inhibitor when the score in the sample surpasses a threshold.

All the above methods according to the invention comprise measuring the level of the biomarkers of the invention (ASCL1 and SOX2) in a sample and assessing said biomarker levels or a derived score (based on said levels) versus a threshold. Typically, each biomarker (i.e. ASCL1 and SOX2) has its threshold (which can be established as described elsewhere herein) and the (measured) levels of ASCL1 and SOX2 are each assessed versus its respective threshold, wherein the patient is identified as more likely to respond to a treatment comprising a KDM1A inhibitor (or, as applicable, the patient is identified as one who may benefit from a treatment comprising a KDM1A inhibitor, etc) when the level of each of ASCL1 and SOX2 in the sample surpasses its threshold. Alternatively, the (measured) levels of ASCL1 and SOX2 in the sample can be used to generate a score for the sample, using a classification algorithm; in such a case a threshold for the score will apply (which can be established as described elsewhere herein) and the patient is identified as more likely to respond to a treatment comprising a KDM1A inhibitor (or, as applicable, the patient is identified as one who may benefit from a treatment comprising a KDM1 A inhibitor, etc) when the score in the sample surpasses the threshold.

In some embodiments of any of the preceding aspects, the method further comprises a step of obtaining or providing a sample from the patient. The obtaining/providing step is prior to the measuring of the level of the biomarkers (and prior to the administration of any treatment comprising a KD 1 A inhibitor to the patient from which the sample is to be obtained/provided). In some embodiments of any of the preceding aspects, the method further comprises recommending, prescribing or administering a therapeutically effective amount of a treatment comprising a KDM1A inhibitor to the patient if the patient is identified as more likely to respond to a treatment comprising a KDM1A inhibitor.

In some embodiments of any of the preceding aspects, the method can optionally further comprise recommending that the patient is not treated with a KDM1A inhibitor if the patient is identified as less likely to respond to a treatment comprising a KDM1A inhibitor.

In the methods according to the invention the levels of ASCL1 and SOX2 can be determined either as mRNA levels or as protein levels, using any methods known in the art for measuring mRNA or protein levels, including the methods as described herein.

In the methods according to the invention, mRNA from a sample can be directly used for determining the level of the biomarker. In the methods according to the present invention, the level can be determined by hybridization. In the methods according to the present invention, the RNA can be transformed into cDNA (complementary DNA) copy using methods known in the art. Methods for detecting can include but are not limited to quantitative reverse transcriptase polymerase chain reaction (qRT-PCR), gene expression analyses, RNA sequencing, nanopore sequencing, microarray analyses, gene expression chip analyses, (in situ) hybridization techniques, RNAscope, and chromatography as well as any other techniques known in the art, e.g. those described in Ralph Rapley,“The Nucleic Acid Protocols Handbook”, published 2000, ISBN: 978-0-89603-459-4. Methods for detecting RNA can include but are not limited to PCR, real-time PCR, digital PCR, hybridization, microarray analyses, as well as any other techniques known in the art, e.g. those described in Leland et al,“Handbook of Molecular and cellular Methods in Biology and Medicine", published 2011, ISBN 9781420069389.

In the methods according to the invention, the method can comprise detecting the protein expression level of a biomarker. Any suitable methods of protein detection, quantization and comparison can be used, such as those described in John M. Walker,“The Protein Protocols Handbook”, published 2009, ISBN 978-1-59745-198-7. The protein expression level of a biomarker can be determined by immune assays which include the recognition of the protein or protein complex by an antibody or antibody fragment, comprising but not limited to enzyme linked immunosorbent assays (ELISA), "sandwich" immunoassays, immunoradiometric assays, in situ immunoassays, alphaLISA immunoassays, protein proximity assays, proximity ligation assay technology (e.g. protein qPCR), western blot analysis, immunoprecipitation assays, immunofluorescent assays, flow cytometry, immunohistochemistry (IHC), immunoeletrophoresis, protein immunostaining, confocal microscopy; or by similar methods in which the antibody or antibody fragment is substituted by a chemical probe, aptamer, receptor, interacting protein or any other biomolecule recognizing the biomarker protein in a specific manner; or by Forster / fluorescence resonance energy transfer (FRET), differential scanning fluorimetry (DSF), microfluidics, spectrophotometry, mass spectrometry, enzymatic assays, surface plasmon resonance, or combinations thereof. Immunoassays may be homogeneous assays or heterogeneous assays. In a homogeneous assay the immunological reaction usually involves the specific antibody, a labeled analyte, and the sample of interest. The signal arising from the label is modified, directly or indirectly, upon the binding of the antibody to the labeled analyte. Both the immunological reaction and detection of the extent thereof can be carried out in a homogeneous solution. Immunochemical labels which may be employed include free radicals, radioisotopes, fluorescent dyes, enzymes, bacteriophages, or coenzymes. In a heterogeneous assay approach, the reagents are usually the sample, the antibody, and means for producing a detectable signal. The antibody can be immobilized on a support, such as a bead, plate or slide, and contacted with the specimen suspected of containing the antigen in a liquid phase. The support is then separated from the liquid phase and either the support phase or the liquid phase is examined for a detectable signal employing means for producing such signal. The signal is related to the presence of the analyte in the sample. Means for producing a detectable signal include the use of radioactive labels, fluorescent labels, or enzyme labels.

In the methods according to the invention, an antibody to the biomarker of interest can be used. In the methods according to the present invention, a kit for detection can be used. Such antibodies and kits are available from commercial sources such as EMD Millipore, R&D Systems for biochemical assays, Thermo Scientific Pierce Antibodies, Novus Biologicals, Aviva Systems Biology, Abnova Corporation, AbD Serotec or others. Alternatively, antibodies can also be synthesized by any known method. The term "antibody" as used herein is intended to include monoclonal antibodies, polyclonal antibodies, single chain antibodies and chimeric antibodies. Antibodies can be conjugated to a suitable solid support (e.g., beads such as protein A or protein G agarose, microspheres, plates, slides or wells formed from materials such as latex or polystyrene) in accordance with known techniques, such as passive binding. Antibodies as described herein may likewise be conjugated to detectable labels or groups able to create signals such as radiolabels (e.g., ³⁵S), enzyme labels (e.g., horseradish peroxidase, alkaline phosphatase), fluorescent labels (e.g., fluorescein, Alexa, green fluorescent protein, rhodamine), phthalocyanine containing beads that can release singlet oxygen after irradiation at 680 nM and provoke emission of light after its subsequent absorption by acceptor beads containing Europium or Therbium, and oligonucleotide labels. Labels can generate signal directly or indirectly. Signal generated can include e.g. fluorescence, radioactivity, or luminescence, in accordance with known techniques.

Antibodies can be substituted by alternative protein capture agents with high affinity and selectivity for the protein biomarkers to be analyzed, including aptamers, affimers, or chemoprobes.

In some embodiments, when measuring biomarker protein levels, the level of the biomarker can be assessed in parts of the SCLC tumor cells, e.g. in the nuclei of tumor cells for example in immunofluorescence analysis of biopsies.

The level of the biomarker can be expressed in any form of mRNA expression or protein expression measurement used in the field, and can be raw data or processed data, i.e. raw data subjected to background substraction, normalizations or other corrections or other mathematical operations or transformations typically used in the field. For example, when measured by microarray hybridization, biomarker levels can be represented by the hybridization signal intensity values of the sample, the Log2(hybridization intensity value of the sample), or the Log2(hybridization signal intensity value of the sample / hybridization signal intensity values of a reference sample). An example of a suitable reference sample is a patient derived tumor sample with high expression level of ASCL1 and SOX2, obtained from a xenograft or PDX model, or a SCLC cell pellet. Hybridization signal values can be obtained using single color or 2 color hybridizations. For example, when measured by qRT-PCR, biomarker levels can be represented by the crossing point-PCR-cycle (Cp) value (which represents the number of cycles needed for the amplification-associated fluorescence to reach a specific threshold level of detection), by the AC_P = C_p- C_p, reference gene, by the 2 ^CP value or 2 ^ACP value. For example, when measured by RNA sequencing, biomarker levels can be represented by the Reads Per Million (RPM), Reads Per Kilobase Million (RPKM), Fragments Per Kilobase Million (FPKM),or Transcripts Per Million (TPM) values. For example, when measured by Western blot, biomarker levels can be represented as integrated density (A.U) of the corresponding bands after image analysis, either as raw integrated density or normalized by protein content and/or as a ratio relative to a reference sample. For example, when measured by immunostaining, biomarker levels can be represented as integrated density (A.U) of nuclear signals after image analysis, either as raw integrated density (A.U)/area unit (pixel² or pm²), or as integrated density/nucleus, or as a ratio relative to a reference sample. For example, when measured by ELISA, biomarker levels can be represented as R.L.U (relative light units) or Absorbance units, either raw or background-corrected or normalized by total protein content and/or as a ratio relative to a reference sample.

In some embodiments of any of the methods according to the invention, the biomarker level (i.e. the ASCL1 level and the SOX2 level) is an mRNA expression level. Preferably, the mRNA expression level is measured by qRT- PCR .

In some embodiments of any of the methods according to the invention, the biomarker level is a protein expression level. Preferably, the protein expression level is measured by fluorescence immunohistochemistry.

In some embodiments, biomarker expression levels in immunofluorescent stainings are visually classified, based on staining intensity levels, as high, medium, low or undetectable, with values 3, 2, 1 and 0, respectively. Samples with medium and high levels (values 2 and 3) are considered“positive" (i.e. surpassing the threshold for the respective biomarker), whereas samples with undetectable or low levels (values 0 and 1) are considered“negative” (i.e. not surpassing the threshold for the respective biomarker). When a sample is considered “positive” for both biomarkers (i.e. when both ASCL1 and SOX2 levels in the sample are each classified as level 2 or 3), the patient is identified as more likely to respond to a treatment comprising a KDM1A inhibitor/ the patient is identified that he/she may benefit from a treatment comprising a KD 1 A inhibitor.

Alternatively, in some embodiments, biomarker expression levels in immunofluorescent staining images can be quantified. A DNA dye, e.g.. DAPI staining can be used to localize the nuclei using fluorescence quantification. Expression of SOX2 and ASCL1 in the nuclei can be analyzed using immunofluorescence quantification. Individual images from biomarker and DAPI stainings can be analyzed using imaging software, e.g. using ImageJ. Signal may be obtained by background subtraction and normalized to a reference (calibrator) sample. A suitable calibrator sample is a sample that has high and homogeneous nuclear expression of both biomarkers, e.g. an NCI-H1417- derived xenograft sample. Normalized quantification values may be expressed in % or as ratios relative to the calibrator sample. The threshold for each biomarker may be established as a fraction of the signal of the calibrator sample and shall be chosen to be higher than the (mean) signal of the negative control sample(s) used (which can be e.g. normal lung biopsy or xenograft samples that have low or undetectable expression of both biomarkers). Preferentially, the thresholds are at least the mean signal of the negative control samples plus 1 SD, 2 SD or 3 SD, wherein SD means standard deviation.

In some embodiments of the methods and uses according to the invention, the term“when the level of each of ASCL1 and SOX2 in the sample surpasses a threshold” or the like can mean“when the level of each of ASCL1 and SOX2 in the sample is increased in comparison to a control”. In this context, when the level of each of ASCL1 and SOX2 in the sample is increased in comparison to a control, the patient is identified as more likely to respond to a treatment comprising a KDM1A inhibitor / the patient is identified that he/she may benefit from a treatment comprising a KDM1A inhibitor. The terms“control” or“reference” are used interchangeably herein. A non-limiting example of a“control” (e.g. a“control value”) or“reference” (e.g. a“reference value) may be the level of ASCL1 and SOX2, respectively, in a sample or pool of samples from one or more healthy individual(s)/subject(s). A healthy individual/subject may, for example, be an individual/subject that is not suffering from SCLC as defined herein, particularly one that is not suffering from SCLC at the time of obtaining the sample from the individual/subject. Alternative, a healthy individual/subject may, for example, be an individual/subject that is not suffering from a disease or disorder associated with increased levels of each of ASCL1 and SOX2. Preferably, the healthy individual/subject is a human. Another non-limiting example of a“control” (e.g. a“control value”) or“reference” (e.g. a“reference value) may be the level of ASCL1 and SOX2, respectively, in a sample or pool of samples from a“non-responder”, e.g. one or more patients that suffer from SCLC and are known to be not responsive to a KDM1A inhibitor. Another example for a“non-responder” control is(are) (a) cell line(s)/cell(s)/tissue(s) that show(s) no response to a KDM1A inhibitor in an in vitro, ex-vivo or (patient derived) xenograft test. Another non-limiting example of a“control” is an“internal standard”, for example purified or synthetically produced proteins and/or peptides or a mixture thereof, or corresponding nucleic acids, where the amount of each protein/peptide (or corresponding nucleic acid) is gauged by using the“non-responder” control described above. In particular, this“internal standard” can contain the protein(s) (or a corresponding nucleic acid) ASCL1 and SOX2 as described and defined herein. A non-limiting example of a “control" (e.g. a“control value”) or“reference" (e.g. a“reference value) may be the level of ASCL1 and SOX2, respectively, in a sample from the patient to be identified herein, if, for example, the sample was obtained before the patient suffered SCLC, before the patient was prone to (or at risk of) suffering from SCLC cancer, or if the sample was obtained when the patient had (fully) recovered from a previous SCLC.

In some embodiments of any of the preceding aspects, the sample is a SCLC biopsy, preferably a SCLC biopsy rich in SCLC cells.

Preferably, in any of the methods of the invention, the patient is a human patient.

It is preferred herein that the above (diagnostic) methods are in vitro methods.“In vitro”, as used herein, means that the methods of the invention as described above, such as methods of identifying patients having SCLC that are more likely to respond to treatment comprising a KDM1A inhibitor and the like, are not performed in vivo, i.e. directly on a patient, but outside of a living human (or other mammal), on a sample obtained from and separated/isolated from the patient (i.e. removed from its in vivo location). In a further aspect, the invention provides a use of ASCL1 and SOX2 in a method of identifying a patient having SCLC who is more likely to respond to a treatment comprising a KDM1A inhibitor.

In a further aspect, the invention provides a use of ASCL1 and SOX2 in a method of assessing the likelihood of response of a patient having SCLC to a treatment comprising a KDM1 A inhibitor.

In another aspect, the invention provides a kit for identifying a patient having SCLC who is more likely to respond to a treatment comprising a KDM1A inhibitor, the kit comprising one or more agents for measuring the level of ASCL1 and SOX2 in a sample, and optionally, instructions for use.

By using the above methods, it is possible to identify a subgroup of patients having SCLC that have a higher chance of response, i.e. are more likely to respond to or to benefit from receiving treatment comprising a KDM1AL The present invention also relates to therapeutic methods for treating those SCLC patients so identified.

Accordingly, in a further aspect, the invention provides a method of treating a patient having SCLC, the method comprising administering to the patient a therapeutically effective amount of a treatment comprising a KDM1A inhibitor, if the patient has been identified as more likely to respond to a treatment comprising a KDM1A inhibitor using a method according to any of the preceding aspects prior to initiating the treatment comprising a KDM1A inhibitor.

In a further aspect, the invention features a method of treating a patient having SCLC, the method comprising measuring the level of ASCL1 and SOX2 in a sample from the patient prior to initiating the treatment, identifying the patient as more likely to respond to a treatment comprising a KDM1A inhibitor when the level of each of ASCL1 and SOX2 in the sample surpasses a threshold, and administering a therapeutically effective amount of a treatment comprising a KDM1 A inhibitor to the patient if identified as more likely to respond.

In some embodiments of any of the preceding aspects, the method further comprises a step of obtaining or providing a sample from the patient.

Preferably, in any of the therapeutic methods and uses according to the invention, the patient is a human patient.

The disclosure elsewhere regarding methods of measuring biomarker levels, sample types and the like are likewise applicable to the above methods of treatment and related therapeutic uses.

KDM1A inhibitors which can be used in accordance with the present invention include any KDM1A inhibitor currently known in the art or that may be reported in the future. Preferably, the KDM1A inhibitor is a small molecule. Both irreversible and reversible KDM1A inhibitors have been reported. Irreversible KDM1A inhibitors exert their inhibitory activity by becoming covalently bound to the FAD cofactor within the KDM1A active site and are typically based on a 2-cyclyl-cyclopropylamino moiety such as a 2-(hetero)arylcyclopropylamino moiety. Reversible inhibitors of KDM1A have also been disclosed. Preferably, the KDM1A inhibitor should be active in cells. Cellular activity of KDM1 A inhibitors can be determined for example using well-established in vitro cellular assays for KDM1 A inhibitors, like for example an SCLC cell viability assay (such as e.g. the assays as described in Example 1 herein or in Mohammad et al, 2015, supra) or an acute myeloid leukemia cell line differentiation assay (such as e.g. the assay as described in Lynch et al, Anal Biochem. 2013 Nov 1 ;442(1):104-6. doi: 10.1016/j.ab.2013.07.032).

Examples of KDM1A inhibitors which can be used in accordance with the present invention include, but are not limited, to those disclosed in: WO2010/043721 , WO2010/084160, WO2011/035941 , WO2011/042217, WO2011/131697, WO2012/013727, WO2012/013728, WO2012/045883, WO2013/057320, WO2013/057322, WO2010/143582, WO2011/022489, WO2011/131576, WO2012/034116, WO2012/135113, WO2013/022047, WO2013/025805, WO2014/058071 , WO2014/084298, W02014/086790, WO2014/164867, WO2014/205213, WO2015/021128, WO2015/031564, W02007/021839, W02008/127734, WO2014/164867 WO2015/089192, WO2015/123408, WO2015/123424, WO2015/123437, WO2015/123465, WO2015/156417, W02015/181380, WO2016/123387, WO2016/130952, WO2016/172496, WO2016/177656, WO2017/027678, WO2012/071469, WO2013/033688, WO2014/085613, WO2015/120281 , WO2015/134973, WO2015/168466, O2015/200843, W02016/003917, WO2016/004105, WO2016/007722, WO2016/007727, WO2016/007731 , WO2016/007736, WO2016/034946, WO2016/037005, WO2016/161282, WO2016172496 WO2017/004519, WO2017/027678, WO2017/079476, WO2017/079670, WO2017/090756, WO2017/109061 , WO2017/116558, WO2017/114497, WO2017/149463, WO2017/157322, WO2017/195216, WO2017/198780, WO2017/215464, WO2018/081342, WO2018/081343, US2010-0324147, US2015-0065434, US2017-0283397, CN106045862, CN104119280, CN103961340, CN103893163, CN103319466, CN103054869, CN105985265, CN106432248, CN106478639, CN106831489, CN106928235, CN107033148 CN107174584, CN107176927, CN107474011 , CN107501169, and CN 107936022, as well as

including any optically active stereoisomer thereof, or any pharmaceutically acceptable salt thereof.

A particularly preferred KDM1A inhibitor is (trans)-N1-((1 R,2S)-2-phenylcyclopropyl)cyclohexane-1 ,4-diamine [CAS Reg, No. 1431304-21-0] or a pharmaceutically acceptable salt thereof, more preferably, (trans)-N1-((1 R,2S)-2- phenylcyclopropyl)cyclohexane-1 , 4-diamine bis-hydrochloride [CAS Reg. No. 1431303-72-8], The compound (trans)- N1-((1 R,2S)-2-phenylcyclopropyl)cyclohexane-1 , 4-diamine, which can also be named as (1 r,4S)-N1-((1 R,2S)-2- phenylcyclopropyl)cyclohexane-1 , 4-diamine, is known as ORY-1001 or iadademstat, and has the chemical structure depicted below:

ORY-1001 has been disclosed for example in WO2013/057322, see example 5 therein. Pharmaceutical formulations comprising ORY-1001 for administration to patients can be prepared following methods known to those skilled in the art, for example as described in WO2013/057322.

KDM1A inhibitors may be administered as the sole API, i.e. as monotherapy, or may be administered in combination of one or more additional APIs, such as other anticancer agents used for the treatment of SCLC. While it is possible that a KDM1A inhibitor (or a treatment comprising a KDM1A inhibitor) may be administered for use in therapy directly as such, it is typically administered in the form of a pharmaceutical composition, which comprises the compound as active pharmaceutical ingredient together with one or more pharmaceutically acceptable excipients or carriers. Any reference to the KDM1A inhibitor herein includes a reference to the compound as such, i.e. the corresponding compound in non-salt form (e.g., as a free base) or in the form of any pharmaceutically acceptable salt or solvate thereof, as well as a reference to a pharmaceutical composition comprising said compound (or a pharmaceutically acceptable salt or solvate thereof) and one or more pharmaceutically acceptable excipients or carriers.

The KDM1A inhibitor may be administered by any means that accomplish the intended purpose. Examples include administration by the oral or parenteral (including e.g. intravenous or subcutaneous) routes.

For oral delivery, the compound can be incorporated into a formulation that includes pharmaceutically acceptable carriers such as binders (e.g., gelatin, cellulose, gum tragacanth), excipients (e.g., starch, lactose), lubricants (e.g., magnesium stearate, silicon dioxide), disintegrating agents (e.g., alginate, Primogel, and corn starch), and sweetening or flavoring agents (e.g., glucose, sucrose, saccharin, methyl salicylate, and peppermint). The formulation can be orally delivered, e.g., in the form of enclosed gelatin capsules or compressed tablets. Capsules and tablets can be prepared by any conventional techniques. The capsules and tablets can also be coated with various coatings known in the art to modify the flavors, tastes, colors, and shapes of the capsules and tablets. In addition, liquid carriers such as fatty oil can also be included in capsules.

Suitable oral formulations can also be in the form of suspension, syrup, chewing gum, wafer, elixir, and the like. If desired, conventional agents for modifying flavors, tastes, colors, and shapes of the special forms can also be included. In addition, for convenient administration by enteral feeding tube in patients unable to swallow, the active compounds can be dissolved in an acceptable lipophilic vegetable oil vehicle such as olive oil, corn oil and safflower oil.

The compound can also be administered parenterally in the form of solution or suspension, or in lyophilized form capable of conversion into a solution or suspension form before use. In such formulations, diluents or pharmaceutically acceptable carriers such as sterile water and physiological saline buffer can be used. Other conventional solvents, pH buffers, stabilizers, anti-bacteria agents, surfactants, and antioxidants can all be included. For example, useful components include sodium chloride, acetates, citrates or phosphates buffers, glycerin, dextrose, fixed oils, methyl parabens, polyethylene glycol, propylene glycol, sodium bisulfate, benzyl alcohol, ascorbic acid, and the like. The parenteral formulations can be stored in any conventional containers such as vials and ampoules.

The pharmaceutical compositions, like oral and parenteral compositions, can be formulated in unit dosage forms for ease of administration and uniformity of dosage. As used herein,“unit dosage forms” refers to physically discrete units suitable as unitary dosages for administration to subjects, each unit containing a predetermined quantity of active ingredient calculated to produce the desired therapeutic effect, in association with one or more suitable pharmaceutical carriers. In therapeutic applications, pharmaceutical compositions are to be administered in a manner appropriate to the disease to be treated, as determined by a person skilled in the medical arts. An appropriate dose and suitable duration and frequency of administration will be determined by such factors as the condition of the patient, the severity of the disease, the specific KDM1 A inhibitor administered, the method of administration, and the judgement of the attending physician, among others. In general, an appropriate dose and administration regimen provides the pharmaceutical composition in an amount sufficient to provide therapeutic benefit, for example an improved clinical outcome, such as more frequent complete or partial remissions, or longer disease-free and/or overall survival, or lessening of symptoms severity, or any other objetively identifiable improvement as noted by the clinician. Effective doses may generally be assessed or extrapolated using experimental models like dose-response curves derived from in vitro or animal model test systems. The skilled person shall be able to determine suitable dosages and treatment regimens based on the above factors.

The pharmaceutical compositions of the invention can be included in a container, pack or dispenser together with instructions for administration.

Examples

The following examples are provided for illustration of the invention. They should not be considered as limiting the scope of the invention, but merely as being representative thereof.

Example 1 : Classification of sensitive and resistant cell lines to KDM1 A inhibitors

For biomarkers identification, seven SCLC cell lines were classified for their response to KDM1A inhibitor treatment based on the results of viability assays performed after either 4 or 10 days of treatment with the KDM1A inhibitor, ORY-1001 (as described herein elsewhere). For 4 days viability assays, SCLC cell lines were seeded in 384-well plates in a final volume of 40mί of the optimized medium recommended by the provider of the each cell line, supplemented with ORY-1001 (maximum concentration: 50mM; 18 serial 1 :2 dilutions tested). After 4 days incubation at 37°C in a 5% C02-controlled atmosphere, cell viability was evaluated using the CellTiterGlo® assay according to manufacturer protocol (Promega). EC50 values were calculated using the the Microsoft Excel software to normalize against untreated cells (100% growth) and no CellTiterGlo® reagent (100% growth inhibition). Quantification of viability after 10 days exposure to ORY-1001 was performed in a 96-well plate format (maximum concentration 1 mM, 100nM for the NCI-H187 cell line). Cells were initially seeded in 100mί of medium (RPMI-1640 10% FBS 2mM glutamine, with the exception of the NCI-H1876 cell line that was cultured in HITES medium, prepared adding 4.5mM glutamine, 0.005mg/mL Insulin, 0.01 mg/mL Transferrin, 30nM Sodium Selenite, 10nM Hydrocortisone, 10nM beta- estradiol to DMEM:F12 5%FBS) and incubated at 37°C in a 5% C02-controlled atmosphere. On day 6, additional 100mί of ORY-1001 -containing medium were added. After 4 additional days, residual viability was measured using the Alamar Blue® assay (Thermo Fisher Scientific). After background subtraction, normalization was performed against vehicle treated cells. EC50 values were calculated after fitting non-linear model using the GraphPad Prism software.

Cell lines were classified as sensitive to KDM1A inhibition when growth inhibition following treatment with ORY-1001 was equal or more than 25% and the EC50 was below 10 nM. SHP77 was classified as sensitive in part, because while not sensitive under the conditions tested here, it has been reported to be sensitive to ORY-1001 in other assays. The results obtained are shown in Table 1 , below. Reported in the table is the concentration achieving a growth reduction of 50% (EC50) and the maximal percentage of growth inhibition (Max Response%) achieved at the highest dose tested. Table 1 :

Example 2. Identification of ASCL1 and SOX2 as biomarkers of responsiveness to KDM1A inhibitors

In order to identify genes that can discriminate SCLC cell lines sensitive and resistant to KDM1A inhibitor treatment, RNASeq analysis was performed on four SCLC cell lines that were identified as KDMIAi sensitive, one KDMIAi partially sensitive and two KDMIAi resistant. Details on their responsiveness to KDMIAi and classification are provided in Example 1 above.

Cell pellets for gene expression analysis

Cells were grown in flasks for 6 consecutive days. On the final day of the assay, cells were collected in a Falcon tube, counted and then centrifuged for 4 min at 1200 rpm. The supernatant was discarded and cells were suspended in 1 mL of PBS and transferred to an eppendorf samples to be centrifuged for 5 min at 3000 rpm in an eppendorf centrifuge at 4°C. Finally, the supernatant was discarded and the pellets frozen at -80°C.

RNA Sequencing

Total RNA was isolated from specimens using the QIAGEN miRNeasy Mini Kit. RNA isolation was performed using the QIAgen RNeasy Mini Kit. The RNeasy Mini Kit combines selective binding properties of a silica-gel-based membrane with the speed of microspin technology. Briefly, tissue was homogenized and lysed in RLT buffer (with beta-mercaptoethanol) using Lysing Matrix D tubes (MP Biomedicals LLC) or vortexing. Ethanol was added to the homogenate to provide appropriate binding conditions for all RNA molecules longer than 200 nucleotides (nt). The sample was applied to an RNeasy Mini column where the total RNA binds to the membrane and contaminants were efficiently washed away. High-quality RNA was then eluted with nuclease-free water. The resulting RNA was quantitated and integrity assessed using an Agilent Bioanalyzer.

Library generation was done using lllumina TruSeq Stranded mRNA Library Preparation. Cluster generation and sequencing of libraries was performed on the lllumina HiSeq. Total RNA samples were converted into cDNA libraries using the TruSeq Stranded mRNA Sample Prep Kit (lllumina). Starting with 100 ng of total RNA, polyadenylated RNA (primarily mRNA) was selected and purified using oligo-dT conjugated magnetic beads. This mRNA was chemically fragmented and converted into single-stranded cDNA using reverse transcriptase and random hexamer primers, with the addition of Actinomycin D to suppress DNA-dependent synthesis of the second strand. Double-stranded cDNA was created by removing the RNA template and synthesizing the second strand in the presence of dUTP in place of dTTP. A single A base was added to the 3’ end to facilitate ligation of sequencing adapters, which contain a single T base overhang. Adapter-ligated cDNA was amplified by polymerase chain reaction to increase the amount of sequence-ready library. During this amplification the polymerase stalls when it encounters a U base, rendering the second strand a poor template. Accordingly, amplified material used the first strand as a template, thereby preserving the strand information. Final cDNA libraries were analyzed for size distribution and using an Agilent Bioanalyzer (DNA 1000 kit), quantitated by qPCR (KAPA Library Quant Kit), then normalized to 2 nM in preparation for sequencing. The Standard Cluster Generation Kit v5 bound cDNA libraries to the flow cell surface. The cBot isothermally amplified the attached cDNA constructs to create clonal clusters of -1000 copies each. The DNA sequence was directly determined using sequencing-by-synthesis technology via the TruSeq SBS Kit.

The following Quality Control metrics were applied: samples had 100ng of input RNA and had a RIN value ³ 7.0 to move forward with library preparation. A minimum total of 30 Million 50 bp paired-end reads were generated per individual sample. A minimum of 28.5 million reads were delivered after subtracting out various off-target sequences such as ribosomal RNA, phiX, homopolymer repeats, and globin RNA.

The lllumina HiSeq software reports the total number of clusters (DNA fragments) loaded in each lane, percent passing sequencing quality filters (which identifies errors due to overloading and sequencing chemistry), a phred quality score for each base of each sequence read, overall average phred scores for each sequencing cycle, and overall percent error (based on alignment to the reference genome). For each RNA-seq sample, the percentage of reads that contain mitochondrial and ribosomal RNA is calculated. The FASTQC package is used to provide additional QC metrics (base distribution, sequence duplication, over represented sequences, and enriched kmers) and a graphical summary. Raw reads were aligned against the human genome (hg19) using GSNAP and recommended options for RNASeq data. In addition to the genome sequence, GSNAP is given a database of human splice junctions and transcripts based on Ensembl v73. Resulting SAM files are then converted to sorted BAM files using Samtools. Gene expression values were calculated both as RPKM values following Mortazavi et al. (Nat Methods (2008) 5(7):621-8) and as read counts. Normalized read counts were obtained using the R package DESeq2. The data are reported as mean of Log2(RPKM) of three independent experiments. RPKM stands for reads per kilobase per million.

Results

Gene expression of ASCL1 and SOX2 in SCLC cell lines measured by RNASeq is shown in Table 2, below. The table indicates the average (A v.) and standard deviation (SD) in Log2(RPKM). Color code shown is based on the gene expression levels; the darker the color, the higher the expression of the biomarker. For comparison, GAPDH expression is reported in parallel to show stable expression across sample of this reference gene.

Table 2:

These results are graphically represented in Figure 1 , which is a dot plot representing expression of ASCL1 (Y-axis) and SOX2 (X-axis) measured by RNA-seq as described above for SCLC cell lines sensitive, sensitive in part or resistant to KDM1A inhibition. Based on the RNASeq data generated for these cell lines, it was identified that all ORY-1001 sensitive and partially sensitive cell lines express high level of ASCL1 and medium-to-high levels of SOX2, while in SCLC cell lines resistant to ORY-1001 treatment, either ASCL1 or SOX2 were detected at very low levels (Log2(RPKM ) £ 0), as shown in Table 2 and Figure 1. Accordingly, ASCL1 and SOX2 may be used as biomarkers to identify SCLC cells, or subjects having SCLC, that are sensitive (i.e. responsive), or more likely to be sensitive (responsive), to treatment with KDM1 A inhibitors, such as ORY-1001.

Example 3. Validation of ASCL1 and SOX2 using qRT-PCR

The biomarkers of responsiveness to KDM1A inhibitors identified in Example 2, ASCL1 and SOX2, were then validated by Taqman qRT-PCR analysis on the same panel of SCLC cell lines described in Example 1 and 2, including two additional SCLC cell lines, one identified as sensitive (DMS53) and one as sensitive in part (NCIH526) to KDMIAi treatment.

Gene expression analysis by qRT-PCR

Total RNA was extracted using the RNeasy Mini Kit and cDNA obtained according to standard procedures, using High Capacity RNA-to-cDNA Master Mix (ThermoFisher Scientific #4390779). qRT-PCR was performed with LightCycler 480 Probes Master (PNT-L-034; Roche #04887301001 ) and using pre-designed and pre-optimized TaqMan Gene Expression Assays from ThermoFisher Scientific. qRT-PCR was performed in triplicate using the Lightcycler 480 Instrument II (Roche; PNT-L-035). Analysis of Cp value by qRT-PCR was performed in triplicate using the following Taqman primers/probeset: - ASCL1 : Hs04187546_g1 (Life Technologies; amplicon length 81 bp, targeting exon 1-2 boundary, RefSeq NM_004316.3, see SEQ ID No. 1

- S0X2 : Hs01053049_s1 (Life Technologies; amplicon length 91 bp, targeting exon 1-1 boundary, RefSeq NM_003106.3, see SEQ ID No. 3 ) - GAPDH : Hs02758991_g1 (Life Technologies; amplicon length 93 bp, targeting exon 6-7 boundary, RefSeq

NM_001256799.2,)

Results

ASCL1 and S0X2 gene expression in this panel of SCLC cell lines as measured by qRT-PCR is shown in Table 3. The table reports Cp values. Exp.R: Experimental replicate; A v.: Average; n.d.: not detected. Each experimental replicate value is the average of three technical replicates The same RNA quantity per sample was analyzed by qRT-PCR. As a comparison, the average Cp expression of GAPDH reference gene are reported and they ranged from 23 to 26 Cp among all samples (see Table 3).

Table 3:

Table 4 shows the average Cp values of all experimental replicates for ASCL1 and SOX2 in SCLC cell lines measured by qRT-PCR. Color code shown is based on the gene expression levels; the darker the color, the higher the expression of the biomarker.

Table 4:

As shown in tables 3 and 4, expression of ASCL1 or SOX2 in KDM1 Ai resistant cells were either not detected at all (Cp value >40) or had absolute Cp values above 35, indicating very low expression. On the other hand, all KDMIAi sensitive SCLC cell lines expressed both ASCL1 and SOX2, thus confirming the findings using RNASeq data described in Example 2. Among the cell lines sensitive in part, one was confirmed to express high levels of both ASCL1 and SOX2 while the other had very low ASCL1 and SOX2 expression.

These results are also graphically represented in Figure 2, which is a dot plot representing gene expression of ASCL1 (Y-axis) and SOX2 (X-axis) measured by qRT-PCR (absolute Cp values) for the above-identified SCLC cell lines sensitive, sensitive in part or resistant to KDM1A inhibition. Plotted values are means of independent experiments as indicated in Table 3. One of the cell lines exhibits a Cp value above 40 for the expression of SOX2; this is indicated by showing the dot in brackets in Fig 2.

The results described herein thus further confirm that SCLC sensitive to KDM1 A inhibitor treatment exhibit in general high expression of both ASCL1 and SOX2, whereas resistant SCLC have low expression of either one or both of ASCL1 and SOX2. ASCL1 and SOX2 levels may thus be used as predictive biomarkers to identify SCLC cells, or subjects with SCLC, having an increased likelihood of responding to KD 1A inhibitor therapy. Example 4. Evaluation of predictive biomarkers of response to KDM1A inhibitors in an expanded panel of SCLC cell lines

For additional biomarker validation of ASCL1 and SOX2 as biomarkers of responsiveness to KDM1 A inhibition, a larger dataset was built and analysed. SCLC viability assay data obtained using ORY-1001 were integrated with publicly available data on response of SCLC cell lines to two additional KDM1A inhibitors, GSK2879552 and GSK- LSD1 , as described in Mohammad et al., Cancer Cell 2015, 28:57-69 (see in particular Figure S2A-B therein, incorporated herein by reference), and then used to classify a larger set of SCLC cell lines as sensitive or resistant to the treatment with KDM1A inhibitors. This expanded panel of SCLC cell lines is shown in table 5, below. The chemical structures for GSK2879552 and GSK-LSD1 are provided in the description.

Table 5:

wherein:

Sensitive: Maximal Response ³ 25% AND EC50 for ORY-1001 < 10 nM, GSK-LSD1 < 20 nM and GSK2879552 < 1000 nM. Cutoff for EC50 was set according to the KDM1A inhibitory potency of each compound.

mm Resistant: Maximal Response < 25% AND/OR EC50 for ORY-1001 > 10 nM, GSK-LSD1 > 20 nM

and GSK2879552 > 1000 nM.

Lack of available data.

*Cell line sensitive to KDMIAi but a treatment longer than 4 days is needed to observe an effect, as shown by the data with the other two KDM1 A inhibitors, tested at longer times. Certain cell lines were classified as sensitive in part because sensitivity of these cell lines to KDM1 A inhibition depends on the assay and different results have been obtained in other assay conditions. The NCIH526 cell line was classified as sensitive in part because while sensitive under the conditions tested here, it has been reported to be resistant to ORY-1001 in other assays. NCIH2081 was classified as sensitive in part because some response to ORY-1001 was observed in vitro but the maximal growth inhibition was very low (10%). Expression of ASCL1 and SOX2 was then evaluated in these KDMIAi sensitive and resistant cells, using the cell lines transcriptomic dataset curated by the Broad Institute (Cancer Cell Line Encyclopedia; https://portals.broadinstitute.org/ccle; CCLE_Expression_Entrez_2012-09-29.gct.txt). Gene expression of ASCL1 , SOX2, and GAPDH in said SCLC cell lines as in CCLE database (Affymetrix microarray data; RMA value) is shown in Table 6, below p value for two-tailed Student’s t-test was calculated using Microsoft Office Excel.

Table 6:

Responsiveness

to KDMIAi

Cell line ID (experimental) ASCL1 SOX2 GAPDH

ASCL1 and SOX2 were differentially expressed between cell lines sensitive and resistant to treatment with KDMIAi (Table 6). In detail, the average expression of ASCL1 was 12.25 and 7.19 normalized probe intensity units for respectively sensitive and resistant cell lines (p value = 3.0E-05). In the same cell lines, the average normalized probe intensity values were 8.96 (sensitive group) and 5.75 (resistant group) for SOX2 (p value = 2.0E-03). Expression of the GAPDH reference gene was similar across all samples (14.60 average normalized probe intensity unites for sensitive cell lines, 14.56 average normalized probe intensity units for resistant cell lines, p value = 6.1 E- 01).

In agreement with the results described in Examples 2 and 3, seven out of eight SCLC cell lines sensitive to KDMIAi highly expressed both ASCL1 and SOX2, while almost all resistant SCLC cell lines were expressing low levels for either one or both of ASCL1 and SOX2. This is further highlighted in Figure 3, which shows a dot plot representing gene expression measured by microarray Affymetrix analysis (RMA values) of ASCL1 and SOX2 in this extended panel of SCLC cell lines sensitive, sensitive in part, or resistant to KDM1 A inhibition. As shown in Figure 3, a clear enrichment is observed in cell lines sensitive to KDM1A inhibitors among the cell lines having high levels of both ASCL1 and SOX2, and viceversa, a clear enrichment in cell lines resistant to KDM1A inhibitors is present among cell lines having low expression of either one or both of ASCL1 and SOX2.

The results described in Example 4 herein therefore further support the use of these two biomarkers in combination to identify/select subjects that are more likely to be responsive to treatment with a KDM1A inhibitor, like ORY-1001.

Gene expression values for sensitive and resistant cell lines were analyzed for each selected biomarker using a receiver operating characteristic curve (ROC curve) with the GraphPad Prism 5.01 Software (“sensitive in part” cell lines excluded from the analysis) . The ROC curve is created by plotting the true positive rate (TPR) against the false positive rate (FPR) at various threshold settings. Figure 4 presents the ROC curves for the expression of ASCL1 (Figure 4A) and SOX2 (Figure 4B) to discriminate sensitive and resistant SCLC cells to KDM1A inhibitors. In base of the ROC curve for each biomarker, the respective threshold levels for the best trade-off between selectivity and specificity (highest likelihood ratio) are reported. Using ASCL1 biomarker alone with a threshold level ³ 8.935 for the expression dataset in Table 6, KDM1A sensitive and resistant cell lines can be discriminated with a sensitivity of 68.42% and a specificity of 100%. Using SOX2 biomarker alone with a threshold level for the given expression dataset ³ 8.030, KDM1A sensitive and resistant cell line can be discriminated with a sensitivity of 84.21% and a specificity of 87.50%.

Predictions on the response to KDM1A inhibition were then made based on a combination of ASCL1 and SOX2, on the panel of SCLC cell lines described in Table 6 (“sensitive in part” cell lines excluded from the analysis) using different algorithms.

In the first instance, a Boolean Conjunction model classification algorithm was built based on the simultaneous compliance of the biomarkers surpassing the individual thresholds for ASCL1 and SOX2 indicated above (Figure 4A and B); in case the algorithm met the conditions indicated in the first raw of Table 7, below, a score of Ί” was obtained, otherwise a score of“0” was obtained. Cell lines were then classified as more likely to respond to a KDMIAi when the score surpassed the threshold i.e. was > 0 (in this case, equal to 1) and less likely to respond when the score was equal to 0 (see Table 7 below). Finally, we evaluated the performance of the Boolean Conjunction classification algorithm. Sensitivity (also called True Positive Rate; TPR), specificity (also called True Negative Rate; TNR), Positive Predictive Value (PPV, also called Precision) and Negative Predictive Value (NPV) were calculated, together with the geometric mean between TPR and TNR, and PPV and NPV. Sensitivity was calculated as the ratio between true positives and the total number of positives. Similarly, the specificity was obtained as the ratio between the number of true negatives and the total number of negatives. The positive predictive value (PPV; also called precision) and negative predictive values (NPV) were calculated using the formulas:

PPV = TP / (TP + FP)

NPV = TN / (TN + FN)

Where TP stands for number of true positives, TN number of true negatives, FP number of false positives and FN number of false negatives.

Table 7:

ASCL1 >8.935

SOX2>8.030

Responsiveness to Boolean Prediction

KDMIAi conjunction according to

ID (experimental) Score threshold (Score >0)

The ASCL1/SOX2 signature exhibits high sensitivity (88%), specificity (100%), precision (100%) and Negative predictive value (95%).

Alternatively, Supportive Vector Machine (SVM) modeling of the DTREG Predictive Modeling Software was used to develop algorithms that can be used to classify samples using the biomarkers. In particular, two different algorithms with different Kernel function, polynomial and radial basis function (RBF) and the respective parameters (C: Cost parameter; Gamma: Kernell Coefficient) reflected in Table 8, Top , were used to evaluate the combination of ASCL1 and SOX2 as predictive biomarkers of responsiveness to KDM1A inhibitors. Using the score and threshold (function) determined by these algorithms, cell lines were classified as more likely to respond or less likely to respond to KDMIAi as indicated and the performance of the SVM models to classify the samples is shown in Table 8, bottom, reflecting the Specificity, Sensitivity and Confusion Matrix for the SVM models generated to predict sensitivity and resistance using the combination of ASCL1 and SOX2 expression.

Table 8:

Sensitivity & Specificity

Positive: Class = Sensitive;

Negative: Class = Resistant

Confusion Matrix

Using this algorithm, the SVM algorithm employing the ASCL1/SOX2 combination has high sensitivity (88%) and high speficity (³95%) using both the polynomial and the RBF fitting. Alternatively, Linear Regression of the DTREG Predictive Modeling Software was used to develop an algorithm to classify samples as more likely to respond or less likely to respond to KDMIAi in function of the combination of the selected biomarkers on the dataset in Table 6 (Table 9). The algorithm calculates the score for each sample and classifies the cell lines as sensitive or resistant to KDMIAi by comparing the score for each sample with the threshold. The performance of the linear regression model generated based on the combination of ASCL1/SOX2 biomarkers has a sensitivity of 87.5% and specificity of 94.74% to predict sensitivity to KDMIAi (Table 9).

Table 9, below, shows the Parameters, Specificity, Sensitivity and Confusion Matrix for the Linear Regression model generated to predict sensitivity and resistance using the combination of ASCL1 and SOX2 expression. C: Cost parameter; Gamma: Kernell Coefficient for the respective functions.

Table 9:

Overall, the use of different classification algorithms (Boolean Conjunction Model, SVM models and linear model) which combined ASCL1 and SOX2 expression levels confirms that this two-marker signature has optimal performance in predicting SCLC response to KDM1A inhibitors.

Example 5. Correlation of ASCL1 and SOX2 mRNA and protein expression levels by Western Blot (WB) and fluorescent immunohistochemistry (IF)

In order to test the correlation between biomarker mRNA and protein levels, ASCL1 and SOX2 were analyzed by WB in SCLC cell lines with either high, medium or low/undetectable expression of these biomarkers, as well as by fluorescent immunohistochemistry of the same sectioned SCLC cell pellets and SCLC patient-derived xenografts (PDXs) with known mRNA levels.

5.1 SOX2 and ASCL1 analysis in SCLC cell lines bv WB

Human small cell lung cancer (SCLC) cell lines NCI-H146, NCI-H510A, NCI-H446, NCI-H526 were grown at 37°C and 5% CCXOn a humid atmosphere in RPMI medium (Sigma) supplemented with 2mM glutamine and 10% FBS (Sigma). Pellets of five million exponentially growing cells were generated and used for whole protein extraction in RIPA buffer (Sigma) supplemented with protease inhibitors (Sigma) and subsequent determination of ASCL1 (Abeam, ab213151) and SOX2 (Abeam, ab97959) levels by WB in 12% PAGE (Life Technologies). Proteins were transferred using the iBIot System (Life Technologies) and after secondary antibody incubation and washing, blots were developed with ECL Prime (Amersham) and photographed with G:BOX Chemi XRQ (Syngene). Ponceau S staining of transferred blots was used a loading control. Quantification of WB signals was carried out with Image J. Integrated densities for each WB band were normalised by the corresponding total protein integrated densities from Ponceau stainings and made relative to NCI-H146 signal.

Results

ASCL1 and SOX2 protein levels were analyzed by WB in SCLC cell lines with either high, medium or low/undetectable expression of these biomarkers, as determined by qRT-PCR (Example 3) and confirmed with publicly available Affymetrix mRNA expression data from the Cancer Cell Line Encyclopedia (CCLE) (see Example 4). WB as obtained is shown in Figure 5A, and the corresponding quantification of ASCL1 and SOX2 protein levels in NCI-H146, NCI-H510A, NCI-H446 and NCI-H526 cell lines is shown in Figure 5B. Protein expression levels of ASCL1 and SOX2 by WB correlate with their respective mRNA levels, as ASCL1 was not detected in NCI-H446 and neither ASCL1 nor SOX2 were detected in NCI-H526 cells, while their expression was the highest in NCI-H146, confirming both mRNA expression levels and specificity of the antibodies used. The correlation between protein and mRNA (CCLE Affymetrix) levels for SOX2 and ASCL1 is plotted in Figures 6A and 6B, respectively; a good correlation is observed, with R values of 0.8957 for SOX2 and 0.9910 for ASCL1. 5.2 SOX2 and ASCL1 analysis bv fluorescent immunohistochemistry on SCLC cell pellets

10 million exponentially growing cells (NCI-H146, NCI-H510A, NCI-H446 and NCI-H526) were fixed in 10% formalin (Sigma) for 1 hour at room temperature, washed in 1X PBS (Sigma), pelleted, included in 1.3% agarose (Sigma) and subsequently dehydrated and included in paraffin for microtome sectioning, 5mhi sections were placed on Superfrost slides, deparaffinised in HistoChoice clearing agent (Sigma) for 5 min twice and hydrated through a decreading ethanol series (2x 100% 5 min, 90% 1 min, 70% 1 min, 30% 1 min, 2x running water). Sections were then subjected to heat-induced antigen retrieval in boiling pH 6 citrate buffer 1X (Sigma) for 20 min. After 20 min left at room temperature, slides were washed in PBS-Triton X100 0.1% (0.1 % PBS-Tx) and blocked in 5% goat serum in 0,1% PBS-Tx for 1 h at room temperature. After blocking, excess liquid was removed with paper tissue by capillarity and sections incubated with primary antibodies diluted in 1% goat serum in 0,1% PBS-Tx overnight at 4°C (1 :500 dilution for SOX2, Abeam ab97959; 1 :100 for ASCL1 Abeam ab213151) and the corresponding negative control (1% goat serum in 0,1% PBS-Tx only). After 3 washes in 0,1% PBS-Tx of 5 min each, slides were incubated with goat anti-rabbit Alexa Fluor 546 secondary antibody (1 :1500 dilution, Life Technologies A11010) for 1 hour at room temperature protected from light. After 5 washes in 0,1% PBS-Tx of 5 min each, excess liquid was removed by capillarity with paper tissue and samples were mounted in Fluoroshield mounting media supplemented with DAPI (Sigma). DAPI is 4',6-diamidino-2-phenylindole, a fluorescent dye that strongly binds to A-T rich regions in DNA and is used to stain the nuclei. Signals from DAPI-stained nuclei were used to identify and analyze co-localization of nuclear-specific SOX2 and ASCL1 signals (see below). Images were taken in a Zeiss Axio fluorescent microscope, with a coupled AxioCam camera (Zeiss).

Quantification of IF intensities

IF images were processed and quantified with imageJ software. For SOX2 and ASCL1 nuclear-specific signal quantification, a mask enclosing the area covered by nuclei was created from the DAPI staining image and then transferred to the corresponding IF image, such that the integrated density was determined in the selected area defined by nuclei only.

Results

In order to validate the antibodies for use in fluorescent immunohistochemistry and to further confirm the correlation between ASCL1 and SOX2 protein and mRNA levels, the same antibodies used for WB were tested in fluorescent immunohistochemistry to interrogate the levels of these biomarkers in formalin-fixed paraffin-embedded sectioned SCLC cell pellets. Fluorescent immunohistochemistry for SOX2, ASCL1 and negative control are shown in Figure 7 (Fig 7A - SOX2, Fig 7B -ASCL1 , Fig 7C - negative control with secondary antibody only (AF546)). In agreement with the data previously shown, SOX2 levels were high in NCI-H146, medium in NCI-H510A and low in NCI-H446 cell lines, while no nuclear-specific expression was detected in NCI-H526 cells. On the other hand, ASCL1 levels were high in NCI-H146, medium in NCI-H510A and absent/undetectable in NCI-H446 and NCI-H526 cells. No expression was observed in the negative control with secondary antibody only (AF546: Alexa Fluor 546). Based on staining intensity, the following staining intensity levels were established and are employed from now onwards in the examples below: high levels will be defined as“Level 3”, medium levels as“Level 2’’, low levels as“Level 1” and absence of signal as“Level 0”.

Table 10, below, shows the quantification of the nuclear biomarker signal in the immunofluorescence images shown in Figure 7 along with the corresponding intensity levels and RMA values from CCLE Affymetrix in NCI-H146, NCI-H510A, NCI-H446 and NCI-H526 cells. Signals were background-corrected and represented relative to the NCI- H146 signal (equivalent to 100%).

Quantification of nuclear signals revealed a highly statistically significant correlation between ASCL1 and SOX2 protein levels detected by fluorescent immunohistochemistry and the corresponding mRNA levels, for the analyzed cell lines (see Figure 8A - SOX2 and 8B - ASCL1). The corresponding R values are 0.8710 and 0.9594 for SOX2 and ASCL1 , respectively.

In view of the good correlation obtained between ASCL1 and SOX2 mRNA and protein levels as shown in Examples 5.1 and 5.2 above, it was confirmed that ASCL1 and SOX2 can be used as predictive biomarkers of responsiveness to KDM1 A inhibition measuring either their mRNA or protein levels.

5.3 SOX2 and ASCL1 analysis by fluorescent immunohistochemistry on patient-derived SCLC xenograft (PDX) tissue microarravs (TMAs)

In order to further confirm the correlation between SOX2 and ASCL1 mRNA and protein levels in human samples, SOX2 and ASCL1 IF were performed on patient-derived SCLC xenograft tissue microarrays with available SOX2 and ASCL1 RNASeq data. Sections of tissue microarrays containing 44 cases of patient-derived SCLC xenografts were purchased from Molecular Response (now Crown Bioscienses) and their corresponding available RNASeq data downloaded from https://oncoexpress.crownbio.com/OncoExpress/index.aspx.

TMAs were deparaffinised in HistoChoice clearing agent (Sigma) for 5 min twice and hydrated through a decreading ethanol series (2x 100% 5 min, 90% 1 min, 70% 1min, 30% 1 min, 2x running water. Sections were then subjected to heat-induced antigen retrieval in boiling pH 6 citrate buffer 1X (Sigma) for 20 min. After 20 min left at room temperature, slides were washed in PBS-Triton X100 0.1% (0.1 % PBS-Tx) and blocked in 5% goat serum in 0,1% PBS-Tx for 1 h at room temperature. After blocking, excess liquid was removed with paper tissue by capillarity and sections incubated with primary antibodies diluted in 1% goat serum in 0,1% PBS-Tx overnight at 4°C (1 :500 dilution for SOX2, Abeam ab97959; 1 :100 for ASCL1 Abeam ab213151) and the corresponding negative control (1% goat serum in 0,1% PBS-Tx only). After 3 washes in 0,1% PBS-Tx of 5 min each, slides were incubated with goat anti-rabbit Alexa Fluor 546 secondary antibody (1 :1500 dilution, Life Technologies A11010) for 1 hour at room temperature protected from light. After 5 washes in 0,1% PBS-Tx of 5 min each, excess liquid was removed by capillarity with paper tissue and samples were mounted in Fluoroshield mounting media supplemented with DAPI (Sigma). Images were taken in a Zeiss Axio fluorescent microscope, with a coupled AxioCam camera (Zeiss).

Classification of PDX IF intensities

IF stainings were analysed with a Zeiss Axio fluorescent microscope, and tumor areas defined by nuclear morphology. Nuclear-specific signal intensities within tumor areas were visually classified in four levels, as described in Example 5.2:

Level Intensity

3 high

2 medium

1 low

0 none

Based on the intensities obtained in the fluorescent immunohistochemistry performed on SCLC cell pellets and their corresponding known mRNA expression (Example 5.2), samples with expression levels surpassing the individual biomarker thresholds of 1 (i.e. medium to high expression Levels 2 and 3) for both ASCL1 and SOX2, or formulated alternatively, samples with a ASCL1/SOX2 Boolean conjunction score surpassing the threshold (> 0 ) ( i.e. simultaneously complying with both conditions for the individual biomarkers) were classified to be derived from patients more likely to respond to KDMIAi. In Figure 9, samples with score surpassing the threshold are indicated as“Positive” and samples with score not surpassing the threshold are indicated as“Negative”. Statistics

Two-tailed Spearman correlation tests with a 95% confidence interval, between RNASeq (Log2FPKM) and IF (visual score) datasets for SOX2 and ASCL1 were carried out with GraphPad Prism software.

Results

SOX2 and ASCL1 IF were performed on patient-derived SCLC xenograft tissue microarrays, with available SOX2 and ASCL1 RNASeq data. All samples were visually analyzed in a fluorescence microscope, tumor areas defined by nuclear morphology and given intensity levels according to the nuclear signals observed within tumor areas, as explained above. Representative ASCL1 and SOX2 stainings from SCLC PDX TMA are shown for each staining intensitity classification level in Figure 9.

In two independent stainings of two consecutive SCLC PDX TMAs sections (N=35 and N=43, respectively), IF visual scores and RNASeq data showed a highly statistically significant (P<0.0001 ) correlation, with Spearman r values of 0.7535 and 0.7659 for SOX2 (Figures 10A and 10B) and 0.8803 and 0.8989 for ASCL1 (Figure 10C and 10D).

The aforementioned results confirm that SOX2 and ASCL1 protein and mRNA levels are also correlated in patient-derived samples and therefore of ASCL1 and SOX2 can be used in human samples as predictive biomarkers of responsiveness to KDM1 A inhibition measuring either their mRNA or protein levels.

Example 6: ASCL1 and SOX2 can be detected in exosomes

Exosomes are microvesicles present in bodily fluids reflecting in their contents the proteasome, genome and transcriptome of parental cells. Thus, exosomes constitute an excellent minimally invasive tool for quantitative biomarker detection. We therefore tested if the detection of our predictive biomarkers of responsiveness to KDM1A inhibitors, ASCL1 and SOX2, was suitable in methods employing exosome-containing samples.

Exosomes were isolated by precipitation from SCLC cell lines and SOX2 and ASCL1 protein levels determined by WB.

ASCL1 and SOX2 detection in exosomes

20 million NCI-H146, NCI-H510A, NCI-H446 and NCI-H526 SCLC cells were seeded in 20ml RPMI medium (Sigma) supplemented with 2mM glutamine (Sigma) and 10% exosome-free FBS (System Biosciences), and incubated in T75 flasks at 37°C and 5% CO2 in a humid atmosphere. After 48h, 15ml of well-resuspended cells were spun at 2.000xg for 30 min at room temperature, the supernatant transferred to a clean tube and cell pellet kept at - 20°C. 10 ml of the cleared medium was used for exosome precipitation with the Total Exosome Isolation reagent (from cell culture media) (Life Technologies) exactly following manufacturer^'s instructions. Exosome pellets were then resuspended either in 80 mI 1x SDS loading buffer for WB or in 40 mI RIPA buffer (Sigma) supplemented with protease inhibitors for protein extraction. After quantification, 2 volumes of RIPA extracts were mixed with 1 volume of 3x SDS loading buffer, heated at 95°C and kept at -20°C until use in WB.

Exosome isolation method was validated by growing NCI-H510A cells, either in the presence of 5mM of the exosome-releasing inhibitor GW4869 (SelleckChem) or vehicle in the conditions specified above, and exosomes isolated with the Total Exosome Isolation reagent (from cell culture media) (Life Technologies) exactly following manufacturer^'s instructions.

Cell pellets kept at -20°C were used for protein extraction in RIPA buffer (Sigma) supplemented with protease inhibidors. After protein quantification with Protein Assay Dye Reagent (Bio-Rad) 7 pg of total protein previously heated at 95°C in 1X SDS loading buffer were used for WB in 12% PAGE (Life Technologies) along with 15 mI of exosome protein extract, using ASCL1 (Abeam, ab213151), SOX2 (Abeam, ab97959) and the lung cancer exosome- specific CD151 marker (Abeam, ab33315) antibodies. Blots were developed with ECL Prime (Amersham) and photographed with G:BOX Chemi XRQ (Syngene). Ponceau S staining of transferred blots was used a loading control.

Results

In order to investigate the feasibility of ASCL1 and SOX2 detection in exosomes, the microvesicle fractions from NCI-H146, NCI-H510A, NCI-H446 and NCI-H526 SCLC cells were isolated by precipitation, and ASCL1 and SOX2 protein levels were analyzed by WB using the method described in Example 5.1. Results obtained are shown in Figure 11 , demonstrating both that ASCL1 and SOX2 can be detected in exosomes and that their expression is in agreement with the expression in parental cells (see Figure 5). ASCL1 was not detected neither in NCI-H446 nor in NCI-H526-derived exosomes (see Figure 11), but highly detected in NCI-H146 and NCI-H510A cells. In its turn, SOX2 was not detected in NCI-H526-derived exosomes (Figure 11), also in agreement with the expression of SOX2 reported in Figure 5.

Furthermore, ASCL1 , SOX2 and the lung cancer-specific exosome marker CD151 signals were significantly reduced or ablated in the exosomal fraction of NCI-H510A cells treated with 5mM of GW4869 (exosome-release inhibitor) compared to vehicle, while expression of these proteins in parental cells remained unchanged (Figure 12). Therefore, these results demonstrate that exosome-derived ASCL1 and SOX2 signals are specific of this microvesicle fraction obtained by precipitation and validate the exosome isolation method employed.

Overall, the results of this Example 6 confirm that measurement of protein levels of the predictive biomarkers of the invention, ASCL1 and SOX2, can be performed to determine responsiveness to KDM1A inhibitors using exosomes as starting material/sample. The present invention refers to the following nucleotide and amino acid sequences:

The sequences provided herein are available in the NCBI database and can be retrieved from www.ncbi.nlm. nih.gov/sites/entrez?db=gene; Theses sequences also relate to annotated and modified sequences. The present invention also provides techniques and methods wherein homologous sequences, and variants of the concise sequences provided herein are used. Preferably, such“variants” are genetic variants, e.g. splice variants.

Exemplary amino acid sequences and nucleotide sequences of human ASCL1 and SOX2 are shown in SEQ ID NO: 1 to 4 herein below. . Exemplary nucleotide and amino acid sequences of human GAPDH (glyceraldehyde-3- phosphate dehydrogenase), used as control gene in some of the Examples, are shown in SEQ ID NO: 5 and 6.

SEQ ID No, 1 : Nucleotide sequence encoding Homo sapiens Achaete-Scute Family bHLH Transcription Factor 1 (ASCL1), mRNA

NCBI Reference Sequence: NM_004316.3.The coding region ranges from nucleotide 572 to nucleotide 1282 (highlighted in bold). It is understood that the mRNA corresponds to the sequence below (i.e. is identical to that sequence) with the exception that the“t” (thymidine) residue is replaced by a“uracil” (u) residue.

ORIGIN

1 agcactctct cacttctggc cagggaacgt ggaaggcgca ccgacaggga tccggccagg

61 gagggcgagt gaaagaagga aatcagaaag gaagggagtt aacaaaataa taaaaacagc

121 ctgagccacg gctggagaga ccgagacccg gcgcaagaga gcgcagcctt agtaggagag

181 gaacgcgaga cgcggcagag cgcgttcagc actgactttt gctgctgctt ctgctttttt

241 ttttcttaga aacaagaagg cgccagcggc agcctcacac gcgagcgcca cgcgaggctc

301 ccgaagccaa cccgcgaagg gaggagggga gggaggagga ggcggcgtgc agggaggaga

361 aaaagcattt tcactttttt tgctcccact ctaagaagtc tcccggggat tttgtatata

421 ttttttaact tccgtcaggg ctcccgcttc atatttcctt ttctttccct ctctgttcct

481 gcacccaagt tctctctgtg tccccctcgc gggccccgca cctcgcgtcc cggatcgctc

541 tgattccgcg actccttggc cgccgctgcg catggaaagc tctgccaaga tggagagcgg

601 cggcgccggc cagcagcccc agccgcagcc ccagcagccc ttcctgccgc ccgcagcctg

661 tttctttgcc acggccgcag ccgcggcggc cgcagccgcc gcagcggcag cgcagagcgc

721 gcagcagcag cagcagcagc agcagcagca gcagcaggcg ccgcagctga gaccggcggc

781 cgacggccag ccctcagggg gcggtcacaa gtcagcgccc aagcaagtca agcgacagcg

841 ctcgtcttcg cccgaactga tgcgctgcaa acgccggctc aacttcagcg gctttggcta

901 cagcctgccg cagcagcagc cggccgccgt ggcgcgccgc aacgagcgcg agcgcaaccg 961 cgtcaagttg gtcaacctgg gctttgccac ccttcgggag cacgtcccca acggcgcggc

1021 caacaagaag atgagtaagg tggagacact gcgctcggcg gtcgagtaca tccgcgcgct

1081 gcagcagctg ctggacgagc atgacgcggt gagcgccgcc ttccaggcag gcgtcctgtc

1141 gcccaccatc tcccccaact actccaacga cttgaactcc atggccggct cgccggtctc

1201 atcctactcg tcggacgagg gctcttacga cccgctcagc cccgaggagc aggagcttct

1261 cgacttcacc aactggttct gaggggctcg gcctggtcag gccctggtgc gaatggactt

1321 tggaagcagg gtgatcgcac aacctgcatc tttagtgctt tcttgtcagt ggcgttggga

1381 gggggagaaa aggaaaagaa aaaaaaaaga agaagaagaa gaaaagagaa gaagaaaaaa

1441 acgaaaacag tcaaccaacc ccatcgccaa ctaagcgagg catgcctgag agacatggct

1501 ttcagaaaac gggaagcgct cagaacagta tctttgcact ccaatcattc acggagatat

1561 gaagagcaac tgggacctga gtcaatgcgc aaaatgcagc ttgtgtgcaa aagcagtggg

1621 ctcctggcag aagggagcag cacacgcgtt atagtaactc ccatcacctc taacacgcac

1681 agctgaaagt tcttgctcgg gtcccttcac ctcctcgccc tttcttaaag tgcagttctt

1741 agccctctag aaacgagttg gtgtctttcg tctcagtagc ccccacccca ataagctgta

1801 gacattggtt tacagtgaaa ctatgctatt ctcagccctt tgaaactctg cttctcctcc

1861 agggcccgat tcccaaaccc catggcttcc ctcacactgt cttttctacc attttcatta

1921 tagaatgctt ccaatctttt gtgaattttt tattataaaa aatctattg tatctatcct

1981 aaccagttcg gggatatatt aagatatttt tgtacataag agagaaagag agagaaaaat

2041 ttatagaagt tttgtacaaa tggtttaaaa tgtgtatatc ttgatacttt aacatgtaat

2101 gctattacct ctgcatattt tagatgtgta gttcacctta caactgcaat tttccctatg

2161 tggttttgta aagaactctc ctcataggtg agatcaagag gccaccagtt gtacttcagc

2221 accaatgtgt cttactttat agaaatgttg ttaatgtatt aatgatgtta ttaaatactg

2281 ttcaagaaga acaaagtta tgcagctact gtccaaactc aaagtggcag ccagttggtt

2341 tgataggtt gccttttgga gatttctatt actgcctttt tttttcttac tgttttatta

2401 caaacttaca aaaatatgta taaccctgtt ttatacaaac tagtttcgta ataaaacttt

2461 ttcctttttt taaaatgaaa ataaaaaaaa //

SEQ ID No. 2: Amino acid sequence of Homo sapiens Achaete-Scute Family bHLH Transcription Factor 1 (ASCL1), protein

UniProtKB/Swiss-Prot: ASCL1J-IUMAN, P50553

MESSAKMESGGAGQQPQPQPQQPFLPPAACFFATAAAAAAAAAAAAAQSAQQQQQQQQQQ

QQAPQLRPAADGQPSGGGHKSAPKQVKRQRSSSPELMRCKRRLNFSGFGYSLPQQQPAAV

ARRNERERNRVKLVNLGFATLREHVPNGAANKKMSKVETLRSAVEYIRALQQLLDEHDAV

SAAFQAGVLSPTISPNYSNDLNSMAGSPVSSYSSDEGSYDPLSPEEQELLDFTNWF SEQ ID No. 3: Nucleotide sequence encoding Homo sapiens SRY-box 2 (SOX2), mRNA NCBI Reference Sequence: NM_003106.3.The coding region ranges from nucleotide 438 to nucleotide 1391 (highlighted in bold). It is understood that the mRNA corresponds to the sequence below (i.e. is identical to that sequence) with the exception that the“t” (thymidine) residue is replaced by a“uracil” (u) residue.

ORIGIN

1 ggatggttgt ctattaactt gttcaaaaaa gtatcaggag ttgtcaaggc agagaagaga

61 gtgtttgcaa aagggggaaa gtagtttgct gcctctttaa gactaggact gagagaaaga

121 agaggagaga gaaagaaagg gagagaagtt tgagccccag gcttaagcct ttccaaaaaa

181 taataataac aatcatcggc ggcggcagga tcggccagag gaggagggaa gcgctttttf

241 tgatcctgat tccagtttgc ctctctcttt ttttccccca aattattctt cgcctgattt

301 tcctcgcgga gccctgcgct cccgacaccc ccgcccgcct cccctcctcc tctccccccg

361 cccgcgggcc ccccaaagtc ccggccgggc cgagggtcgg cggccgccgg cgggccgggc

421 ccgcgcacag cgcccgcatg tacaacatga tggagacgga gctgaagccg ccgggcccgc

481 agcaaacttc ggggggcggc ggcggcaact ccaccgcggc ggcggccggc ggcaaccaga

541 aaaacagccc ggaccgcgtc aagcggccca tgaatgcctt catggtgtgg tcccgcgggc

601 agcggcgcaa gatggcccag gagaacccca agatgcacaa ctcggagatc agcaagcgcc

661 tgggcgccga gtggaaactt ttgtcggaga cggagaagcg gccgttcatc gacgaggcta

721 agcggctgcg agcgctgcac atgaaggagc acccggatta taaataccgg ccccggcgga

781 aaaccaagac gctcatgaag aaggataagt acacgctgcc cggcgggctg ctggcccccg

841 gcggcaatag catggcgagc ggggtcgggg tgggcgccgg cctgggcgcg ggcgtgaacc

901 agcgcatgga cagttacgcg cacatgaacg gctggagcaa cggcagctac agcatgatgc

961 aggaccagct gggctacccg cagcacccgg gcctcaatgc gcacggcgca gcgcagatgc

1021 agcccatgca ccgctacgac gtgagcgccc tgcagtacaa ctccatgacc agctcgcaga

1081 cctacatgaa cggctcgccc acctacagca tgtcctactc gcagcagggc acccctggca

1141 tggctcttgg ctccatgggt tcggtggtca agtccgaggc cagctccagc ccccctgtgg

1201 ttacctcttc ctcccactcc agggcgccct gccaggccgg ggacctccgg gacatgatca

1261 gcatgtatct ccccggcgcc gaggtgccgg aacccgccgc ccccagcaga cttcacatgt

1321 cccagcacta ccagagcggc ccggtgcccg gcacggccat taacggcaca ctgcccctct

1381 cacacatgtg agggccggac agcgaactgg aggggggaga aattttcaaa gaaaaacgag

1441 ggaaatggga ggggtgcaaa agaggagagt aagaaacagc atggagaaaa cccggtacgc

1501 tcaaaaagaa aaaggaaaaa aaaaaatccc atcacccaca gcaaatgaca gctgcaaaag

1561 agaacaccaa tcccatccac actcacgcaa aaaccgcgat gccgacaaga aaacttttat

1621 gagagagatc ctggacttct ttttggggga ctatttttgt acagagaaaa cctggggagg

1681 gtggggaggg cgggggaatg gaccttgtat agatctggag gaaagaaagc tacgaaaaac 1741 tttttaaaag ttctagtggt acggtaggag ctttgcagga agtttgcaaa agtctttacc

1801 aataatattt agagctagtc tccaagcgac gaaaaaaatg ttttaatatt tgcaagcaac

1861 ttttgtacag tatttatcga gataaacatg gcaatcaaaa tgtccattgt ttataagctg

1921 agaatttgcc aatatttttc aaggagaggc ttcttgctga attttgattc tgcagctgaa

1981 atttaggaca gttgcaaacg tgaaaagaag aaaattattc aaatttggac attttaattg

2041 tttaaaaatt gtacaaaagg aaaaaattag aataagtact ggcgaaccat ctctgtggtc

2101 ttgtttaaaa agggcaaaag ttttagactg tactaaattt tataacttac tgttaaaagc

2161 aaaaatggcc atgcaggttg acaccgttgg taatttataa tagcttttgt tcgatcccaa

2221 ctttccattt tgttcagata aaaaaaacca tgaaattact gtgtttgaaa tattttctta

2281 tggtttgtaa tatttctgta aatttattgt gatattttaa ggttttcccc cctttatttt

2341 ccgtagttgt attttaaaag attcggctct gtattatttg aatcagtctg ccgagaatcc

2401 atgtatatat ttgaactaat atcatcctta taacaggtac attttcaact taagttttta

2461 ctccattatg cacagtttga gataaataaa tttttgaaat atggacactg aaaaaaaaaa //

SEQ ID No. 4: Amino acid sequence of Homo sapiens Homo sapiens SRY-box 2 (S0X2), protein UniProtKB/Swiss-Prot: S0X2_HUMAN, P48431

MYNMMETELKPPGPQQTSGGGGGNSTAAAAGGNQKNSPDRVKRPMNAFMVWSRGQRRKMA

QENPKMHNSEISKRLGAEWKLLSETEKRPFIDEAKRLRALHMKEHPDYKYRPRRKTKTLM

KKDKYTLPGGLLAPGGNSMASGVGVGAGLGAGVNQRMDSYAHMNGWSNGSYSMMQDQLGY

PQHPGLNAHGAAQMQPMHRYDVSALQYNSMTSSQTYMNGSPTYSMSYSQQGTPGMALGSM

GSVVKSEASSSPPVVTSSSHSRAPCQAGDLRDMISMYLPGAEVPEPAAPSRLHMSQHYQS

GPVPGTAINGTLPLSHM

SEQ ID No. 5: Homo sapiens glyceraldehyde-3-phosphate dehydrogenase (GAPDH), mRNA

NCBI Reference Sequence: NM_002046.6.The coding region ranges from nucleotide 77 to nucleotide 1084 (highlighted in bold). It is understood that the mRNA corresponds to the sequence below (i.e. is identical to that sequence) with the exception that the T (thymidine) residue is replaced by a“uracil” (u) residue.

ORIGIN

1 gctctctgct cctcctgttc gacagtcagc cgcatcttct tttgcgtcgc cagccgagcc

61 acatcgctca gacaccatgg ggaaggtgaa ggtcggagtc aacggatttg gtcgtattgg

121 gcgcctggtc accagggctg cttttaactc tggtaaagtg gatattgttg ccatcaatga

181 ccccttcatt gacctcaact acatggttta catgttccaa tatgattcca cccatggcaa

241 attccatggc accgtcaagg ctgagaacgg gaagcttgtc atcaatggaa atcccatcac 301 catcttccag gagcgagatc cctccaaaat caagtggggc gatgctggcg ctgagtacgt

361 cgtggagtcc actggcgtct tcaccaccat ggagaaggct ggggctcatt tgcagggggg

421 agccaaaagg gtcatcatct ctgccccctc tgctgatgcc cccatgttcg tcatgggtgt

481 gaaccatgag aagtatgaca acagcctcaa gatcatcagc aatgcctcct gcaccaccaa

541 ctgcttagca cccctggcca aggtcatcca tgacaacttt ggtatcgtgg aaggactcat

601 gaccacagtc catgccatca ctgccaccca gaagactgtg gatggcccct ccgggaaact

661 gtggcgtgat ggccgcgggg ctctccagaa catcatccct gcctctactg gcgctgccaa

721 ggctgtgggc aaggtcatcc ctgagctgaa cgggaagctc actggcatgg ccttccgtgt

781 ccccactgcc aacgtgtcag tggtggacct gacctgccgt ctagaaaaac ctgccaaata

841 tgatgacatc aagaaggtgg tgaagcaggc gtcggagggc cccctcaagg gcatcctggg

901 ctacactgag caccaggtgg tctcctctga cttcaacagc gacacccact cctccacctt

961 tgacgctggg gctggcattg ccctcaacga ccactttgtc aagctcattt cctggtatga

1021 caacgaattt ggctacagca acagggtggt ggacctcatg gcccacatgg cctccaagga

1081 gtaagacccc tggaccacca gccccagcaa gagcacaaga ggaagagaga gaccctcact

1141 gctggggagt ccctgccaca ctcagtcccc caccacactg aatctcccct cctcacagtt

1201 gccatgtaga ccccttgaag aggggagggg cctagggagc cgcaccttgt catgtaccat

1261 caataaagta ccctgtgctc aaccagttaa aaaaaaaaaa aaaaaaaaa II

SEQ ID No. 6: Amino acid sequence of Homo sapiens glyceraldehyde-3-phosphate dehydrogenase (GAPDH), protein UniProtKB/Swiss-Prot: G3P_HUMAN, P04406

MGKVKVGVNGFGRIGRLVTRAAFNSGKVDIVAINDPFIDLNYMVYMFQYDSTHGKFHGTV

KAENGKLVINGNPITIFQERDPSKIKWGDAGAEYWESTGVFTTMEKAGAHLQGGAKRVI

ISAPSADAPMFVMGVNHEKYDNSLKIISNASCTTNCLAPLAKVIHDNFGIVEGLMTTVHA

ITATQKTVDGPSGKLWRDGRGALQNIIPASTGAAKAVGKVIPELNGKLTGMAFRVPTANV

SVVDLTCRLEKPAKYDDIKKVVKQASEGPLKGILGYTEHQVVSSDFNSDTHSSTFDAGAG

lALNDHFVKLlSWYDNEFGYSNRVVDLMAHMASKE

Claims

1. A method of identifying a patient having SCLC who is more likely to respond to a treatment comprising a KDM1A inhibitor, the method comprising measuring the level of ASCL1 and SOX2 in a sample from the patient prior to initiating the treatment comprising a KDM1A inhibitor.

2. The method of claim 1 , further comprising identifying the patient as more likely to respond to a treatment comprising a KDM1A inhibitor when the level of each of ASCL1 and SOX2 in the sample surpasses a threshold.

3. The method of claim 1 , further comprising using the level of ASCL1 and SOX2 in the sample to generate a score for the sample, wherein the patient is identified as more likely to respond to a treatment comprising a KDM1A inhibitor when the score in the sample surpasses a threshold.

4. A method of identifying a patient having SCLC who may benefit from a treatment comprising a KDM1A inhibitor, the method comprising measuring the level of ASCL1 and SOX2 in a sample from the patient prior to initiating the treatment comprising a KDM1A inhibitor.

5. The method of claim 4, further comprising identifying the patient as one who may benefit from a treatment comprising a KDM1A inhibitor when the level of each of ASCL1 and SOX2 in the sample surpasses a threshold.

6. The method of claim 4, further comprising using the level of ASCL1 and SOX2 in the sample to generate a score for the sample, wherein the patient is identified as one who may benefit from a treatment comprising a KDM1A inhibitor when the score in the sample surpasses a threshold.

7. A method of selecting a treatment for a patient having SCLC, the method comprising measuring the level of ASCL1 and SOX2 in a sample from the patient prior to initiating the treatment.

8. The method of any of claims 1 to 7, wherein the ASCL1 level and the SOX2 level is an mRNA expression level.

9. The method of claim 8, wherein the mRNA expression level is measured by qRT-PCR.

10. The method of any of claims 1 to 7, wherein the ASCL1 level and the SOX2 level is a protein expression level.

11. The method of claim 10, wherein the protein expression level is measured by fluorescence immunohistochemistry.

12. The method of any of claims 1 to 11 , wherein the sample is a biopsy.

13. The method of any one of claims 1 to 12, the method further comprising recommending, prescribing or administering a therapeutically effective amount of a treatment comprising a KDM1A inhibitor to the patient if the patient is identified as more likely to respond to a treatment comprising a KDM1A inhibitor.

14. KDM1A inhibitor for use in treating a patient having SCLC, wherein the patient has been identified as more likely to respond to a treatment comprising a KDM1A inhibitor using a method according to any of claims 1 to 12 prior to initiating the treatment comprising a KDM1A inhibitor.

15. A method of treating a patient having SCLC, the method comprising administering to the patient a therapeutically effective amount of a treatment comprising a KDM1A inhibitor, if the patient has been identified as more likely to respond to a treatment comprising a KDM1A inhibitor using a method according to any of claims 1 to 12 prior to initiating the treatment comprising a KDM1A inhibitor.

16. Use of ASCL1 and SOX2 in a method of identifying a patient having SCLC who is more likely to respond to a treatment comprising a KDM1A inhibitor.

17. The method of claims 1 to 13, the KDM1A inhibitor for use of claim 14, the method of treatment of claim 15 or the use of claim 16, wherein the KDM1A inhibitor is (trans)-N1-((1 R,2S)-2-phenylcyclopropyl)cyclohexane-1 ,4- diamine or a pharmaceutically acceptable salt thereof.

18. The method of claims 1 to 13 or 17, the KD 1A inhibitor for use of claim 14 or 17, the method of treatment of claim 15 or 17 or the use of claim 16 or 17, wherein the patient is a human patient.

19. A kit for assessing the likelihood of response of a patient having SCLC to a treatment comprising a KDM1A inhibitor, the kit comprising one or more agents for measuring the level of ASCL1 and SOX2 in a sample, and optionally, instructions for use.