WO2019215394A1 - Arpp19 as biomarker for haematological cancers - Google Patents

Abstract

The present invention relates to ARPP19 as a standalone predictive biomarker for haematological cancers such as leukaemias, especially acute myeloid leukaemia (AML). The invention provides use of ARPP19 and an in vitro method of predicting relapse of haematological cancer in a subject on the basis of the expression level of ARPP19 in a sample obtained from said subject. Also other methods involving ARPP19 as a biomarker are provided, as well as a kit for use in said methods.

Description

ARPP19 AS BIOMARKER FOR HAEMATOLOGICAL CANCERS

F1ELD OF THE INVENTION

The present invention relates to ARPP19 as biomarker for various as pects of haematological cancers such as leukaemias, especially acute myeloid leu kaemia (AML). Thus, the invention provides use of ARPP19 and an in vitro meth od of predicting relapse of haematological cancer in a subject on the basis of the expression level of ARPP19 in a sample obtained from said subject. Also other methods involving ARPP19 as a biomarker are provided, as well as a kit for use in said methods.

BACKGROUND OF THE INVENTION

Acute myeloid leukaemia (AML) is the most common acute leukaemia affecting adults. AML is a heterogeneous clonal hematological malignancy that disrupts normal hematopoiesis and it is one of the most aggressively progressive cancer types. First curative therapies for AML were developed in the 1970s-80s and before this patients with AML had a dismal prognosis of only three months ln the 2010s 50% of the patients with AML under age 60-65 can obtain a full remis sion. Advances in understanding the genetic diversity of AML have enabled identi fication of prognostic subgroups within AML cases. Classification of AML subtypes is also clinically relevant in order to optimize treatment strategies and to reduce treatment related mortality and relapse risk.

Despite advances in our understanding of the molecular biology of AML and although aggressive measures are taken to achieve and maintain remis sion from AML, for many patients relapse is generally expected and prognosis is dismal. Depending on the duration of the antecedent remission, the likelihood of achieving each subsequent remission decreases with each attempt. There are multiple factors that limit overall survival for adults diagnosed with AML: cytoge netic and molecular changes, population heterogeneity, differing response to chemotherapy, and cell cycle state (i.e. leukaemia stem cells) (Roboz & Guzman, 2009). Furthermore the bone marrow microenvironment is a protective niche, which provides resistance to a subset of leukaemia cells to chemotherapy. The difficulty in treating patients with AML is linked to a residual subset of chemo therapy-resistant cells present during remission that eventually drive the leu kaemia relapse.

Novel clinical and molecular markers for AML prognosis and minimal residual disease monitoring are needed. As part of routine clinical diagnostics, AML patient samples are examined with PCR for fusion genes and mutations that are characteristic for AML. However, not all of the genetic abnormalities behind the WHO classification of AML for poor prognosis are known, and actually only for half of the AML patients there is a relevant molecular genetic PCR test availa ble. Thus, there is a clear medical need to identify new genetic changes that could be used to design and optimize treatment strategies for AML patients.

EV11 (ecotropic virus integration site 1) is a haematopoietic sternness and transcription factor with chromatin remodelling activity. Aberrant expression of EV11 gene is a common event in AML and EV11 overexpression has been re ported to be an independent adverse prognostic marker in adult AML. Overex pression of EV11 has been shown to act as an independent adverse prognostic marker for complete remission (CR), overall- (OS), relapse free- (RFS) and event-free (EFS) survival in AML (Hinai & Valk, (2016) Br J Haematol 172, 870- 878).

WT1 (Wilms tumor 1) protein is a transcription factor that plays an important role in cellular development and cell survival. The role of WT1 in the leukemogenesis of AML is generally accepted, but its molecular details are largely unknown. Overexpression of WT1 in AML has been known for two decades al ready. Nowadays it’s thought that WTl-positivity at diagnosis does not have reli able predictive value in adult AML per se, however it is a useful tool for monitor ing minimal residual disease (MRD) in cases where a remission can be achieved (Ujj et al. (2016) Pathol Oncol Res 22, 217-221).

Protein Phosphatase 2A (PP2A) is a trimeric (A-B-C subunits) tumor suppressor complex that plays a critical role in diverse signaling pathways includ ing regulation of the cell cycle, cell proliferation, and neuronal signaling. Deficits in PP2A function are implicated in many human diseases including cancer, Alz heimer’s disease and major depressive disorder ln cancer cells, the tumor sup pressor activity of PP2A is inhibited. However, the PP2A complex proteins are mutated with rather low frequency in all cancer types lnstead, the most prevalent mode of PP2A inhibition in cancer seems to be the overexpression of PP2A inhibi tor proteins, such as C1P2A, SET and ARPP19.

ARPP19 (cAMP-regulated phosphoprotein 19), also known as ARPP- 19, a member of the alpha-endosulfine (ENSA) family, is ubiquitously expressed mitotic PP2A inhibitor that promotes the G2/M transition and the mitotic state. ARPP19 requires phosphorylation of a conserved serine residue (Ser-62) by the Greatwall kinase (mammalian Greatwall orthologue is MAST -like kinase) in order to bind PP2A and subsequently inhibit PP2A activity towards a physiological CDK1 substrate.

W02011/066660 discloses ARPP19 as a member of a leukaemia stem cell signature comprising the expression profile of 80 genes. However, ARPP19 is not disclosed nor suggested as a standalone predictive biomarker for haemato- logical cancers.

BR1EF DESCRIPTION THE INVENTION

An object of the present invention is to provide means and methods of predicting outcome of haematological cancer in a subject. This object is achieved by what is stated in the independent claims. Preferred embodiments of the inven tion are disclosed in the dependent claims.

At least some embodiments of the invention are based on the evalua tion of 80 AML patient samples in order to identify novel molecular markers that accurately define AML patient outcomes upon initial diagnosis and which can be rapidly determined by qPCR. lt was surprisingly found out that higher ARPP19 gene expression alone at diagnosis phase predicts independently of the prognos tic subgroups, for example, relapse of the disease when treated with the standard therapy.

Thus, in one aspect, the invention provides a method of predicting re lapse in a subject having been treated for haematological cancer, wherein the method comprises: assaying the level of ARPP19 expression in samples obtained from the subject at two or more time points; detecting a change in the assayed expression levels of ARPP19 between the two or more time points; and predicting the relapse from the change, wherein increased risk of relapse is determined from an increase in the level of ARPP19 expression, or non-increased risk of relapse is determined from a non-increased level of ARPP19 expression.

ln another aspect, the invention provides use of ARPP19 as a bi omarker for predicting relapse in a subject having been treated for haematologi cal cancer.

ln a further aspect, the invention provides a method of selecting treatment to a subject diagnosed with AML, the method comprising the steps of: determining the expression level of ARPP19 in a biological sample obtained from the subject; determining whether or not the determined expression level of ARPP19 is increased in the sample by comparing said expression level to a con trol; and selecting a treatment regime comprising stem cell therapy if said expres- sion is increased, or selecting a treatment regime comprising chemotherapy if said expression is non-increased or decreased.

ln a still further aspect, the invention provides a kit for use in the pre sent methods set forth above, wherein the kit comprises one or more reagents that specifically detect ARPP19 in a sample obtained from a subject whose hae- matological cancer relapse is to be prognosed, or to whom AML treatment is to be selected.

ln an even further aspect, the invention provides use of a kit compris ing one or more reagents that specifically detect ARPP19 for predicting relapse of haematological cancer in a subject, or for selecting AML treatment to a subject.

Other objectives, aspects, embodiments, details and advantages of the present invention will become apparent from the following figures, detailed de scription, examples, and dependent claims.

BR1EF DESCRIPTION OF THE DRAW1NGS

ln the following the invention will be described in greater detail by means of preferred embodiments with reference to the attached drawings, in which:

Figure 1 shows waterfall blots of analysed genes from the patient co- hortl normalized to GAPDH & b-actin expression and pooled normal bone mar row sample (n=56). On y-axis loglO transformed RQ mRNA expression values. One bar represents one patient. Figure 1A shows that WT1 mRNA expression was highly overexpressed (91%) in diagnosis phase AML patients bone marrow com pared to normal bone marrow. Figure IB shows that EV11 overexpression was 13%, Figure 1C shows that ARPP19 overexpression was 21%.

Figure 2 demonstrates the clinical relevance of a classical risk group ing in AML patients analysed by Kaplan-Meier estimators for overall survival. Shown are ten years survival probabilities. The results shows that in the used pa tient material higher risk group is significantly associated with poor survival of AML patients, p=0.003 by log-rank test. Risk group 1, favorable; Risk group 2, in termediate; Risk group 3, adverse.

Figure 3 shows Kaplan-Meier curves for overall survival and relapse- free survival according to EV11 mRNA expression. Figure 3A demonstrates that high EV11 expression is associated with poor survival of AML patients, p=0.0004 by log-rank test. Figure 3B demonstrates that high EV11 expression is significantly associated with shorter relapse-free survival of AML patients, p<0.0001 by log- rank test.

Figure 4A shows the results of Wilcoxon rank sum test (p=0.03) for ARPP19 mRNA expression levels of patients who had relapse (n=19) during the follow-up time and who did not (n=49). Shown is mean ±SEM. Figure 4B demon strates that higher ARPP19 expression is significantly associated with shorter re- lapse-free survival of AML patients, p=0.029 by log-rank test. Qi= lowest quartile of ARPP19 mRNA expression (n=17). Figure 4C shows that patients in the lowest quartile ARPP19 mRNA expression are assigned to all risk groups, as well as pa tients in the over lowest quartile expressing subgroup (Figure 4D). Figures 4C and 4D demonstrate the risk group independent role of ARPP19 gene expression.

Figure 5A demonstrates ARPP19 mRNA expression in diagnosis (n=9), remission (n=4) and relapse (n=8) samples from patients in the patient cohort2. Shown is mean of replicates ±SEM. **(diagnosis vs. remission) p=0.003, *(remission vs. relapse) p=0.01 by Student’s t-test.

Figure 5B demonstrates ARPP19 mRNA expression in matched diag nosis, remission and relapse samples from three patients with AML in the patient cohort2. *(diagnosis vs. remission) p=0.04, *(remission vs. relapse) p=0.03 by paired Student’s t-test.

Figure 6 shows receiver operating characteristic (ROC) curves for re- lapse-free survival obtained with ARPP19 (Figure 6A), EV11 (Figure 6B) or ARPP19 and EV11 used in combination (Figure 6C).

DETA1LED DESCRIPTION OF THE INVENTION

The present invention is based on identification of ARPP19 as a standalone prognostic biomarker for haematological cancers, especially leukae mias such as AML or a subgroup thereof.

Accordingly, herein is provided use of ARPP19 and an in vitro method of determining or predicting the outcome of haematological cancer in a subject on the basis of the expression level of ARPP19 in a sample obtained from said sub ject. More specifically, said method may be formulated e.g. as a method of deter mining a prognosis of haematological cancer in a subject in need thereof, wherein the method comprises or consists of assaying a sample obtained from said subject for the level of ARPP19 expression, comparing the assayed level of ARPP19 to a control level, and providing a prognosis of said haematological cancer on the ba sis of said comparison. The expression level of ARPP19 correlates with the proba ble outcome of haematological cancer such that increased expression level of ARPP19 as compared with a control expression level in a corresponding healthy sample is indicative of poor outcome, such as reduced overall survival or reduced relapse-free survival. On the other hand, non-increased or decreased expression of ARPP19 predicts for good prognosis.

As used herein, the term "or" has the meaning of both "and"' and "or" (i.e. "and/or"). Furthermore, the meaning of a singular noun includes that of a plural noun and thus a singular term, unless otherwise specified, may also carry the meaning of its plural form ln other words, the term "a" or "an" may mean one or more.

As used herein, the terms "biomarker" and "marker" are interchange able, and refer to a molecule which is differentially present in a sample taken from a subject with haematological cancer as compared to a comparable sample take from a control subject, such as an apparently healthy subject or a subject with the same haematological cancer but having a different prognosis of the dis ease.

As used herein, the term "determining an expression level of ARPP19", and any corresponding expressions, refer to quantifying or semi-quantifying ARPP19 either on the basis of mRNA, cDNA, or protein amount. The term "level" is interchangeable with the terms "amount" and "concentration", and can refer to an absolute or relative quantity. Determining ARPP19 expression may also in clude determining the absence or presence of ARPP19 qualitatively.

To make any prognostic or predictive conclusions on the basis of ARPP19 expression, it should be compared with a relevant control. Once the con trol level is known, the determined or detected ARPP19 level can be compared therewith and the significance of the difference can be assessed using standard statistical methods ln some embodiments, a statistically significant difference be tween the determined ARPP19 level and the control level is indicative of a certain prognosis, such as poor prognosis that may manifest itself e.g. as reduced overall survival or reduced relapse-free survival, or good prognosis that may manifest itself e.g. as increased overall survival or increased relapse-free survival ln some embodiments, a statistically significant difference between the determined ARPP19 level and the control level is indicative of a probable course the disease, such as relapse ln some further embodiments, before being compared with the control, the ARPP19 levels are normalized using standard methods.

ln some embodiments, the expression level of ARPP19 may be deter mined as a relative ratio between the expression level of ARPP19 and that of an appropriate housekeeping gene in the same cells. Non-limiting examples of suita ble housekeeping genes include GAPDH (Glyceraldehyde-3-phosphate dehydro genase), ABL1 (ABL proto-oncogene 1), ACTB (Actin, beta), RRN18S (18S riboso- mal RNA), PGK1 (Phosphoglycerate kinase 1), PP1A (Peptidylprolyl isomerase A (cyclophilin A)), RPL13A (Ribosomal protein L13a), RPLPO (Ribosomal protein, large, P0), B2M (Beta-2-microglobulin), YWHAZ (Tyrosine 3- monooxygenase/tryptophan 5-monooxygenase activation protein, zeta polypep tide), SDHA (Succinate dehydrogenase), TFRC (Transferrin receptor (p90, CD71)), ALAS1 (Amino-levulinate, delta-, synthase 1), GUSB (Glucuronidase, beta), HMBS (Hydroxy-methyl-bilane synthase), HPRT1 (Hypoxanthine phosphoribosyltrans- ferase 1), TBP (TATA box binding protein), and TUBB (Tubulin, beta polypeptide) ln some embodiments, preferred housekeeping genes include GAPDH and ACTB.

As used herein, the term "control" may refer to many different types of controls. For example, "negative control" or "normal control" means the control level of ARPP19, alone or in comparison with another marker or control, deter mined from a non-cancerous tissue of the same subject, or from a sample ob tained from an apparently healthy individual or a pool of apparently healthy indi viduals. "Positive control" may mean, for example, cancerous tissue of the same subject, or a sample obtained from an individual diagnosed with haematological cancer, or a pool of individuals diagnosed with haematological cancer. A prede termined threshold value that is indicative of a given prognosis may also be em ployed as a control, i.e. "threshold control". Accordingly, the threshold value may be a negative control value obtained from a pool of apparently heathy individuals, wherein the value refers to expression levels of ARPP19 in normal, non-cancerous cells; or it may be a positive control value, obtained from a pool of individuals with haematological cancer, wherein the value refers to an expression level of ARPP19 in cancerous tissues ln some embodiments, especially in those relating to monitoring purposes, such as monitoring for a minimal residual disease or for a relapse risk especially in subjects with minimal residual disease, "threshold con trol" refers to a expression level of ARPP19 in a sample obtained from the same subject at a different time point, such as at the time of diagnosis and/or at the time when patient has achieved clinical remission.

"lnternal control" may refer to an expression level of one or more ap propriate housekeeping genes determined from the same cells as the expression level of ARPP19. ln other words, the internal control may mean a loading control which allows quantifying ARPP19 in the sample being analyzed. Examples for an internal control, in other words "loading control" or "quantifying control" include housekeeping genes and housekeeping proteins. The quantity of ARPP19 may al so be measured relative to another biomarker, such as a known AML biomarker (e.g. WT1 or EV11), which in this context may be denoted as "control biomarker". "Control ratio" means the relationship of ARPP19 to any selected control, for ex ample the ratio between ARPP19 and a control biomarker.

ln some preferred embodiments, the control is the expression level of ARPP19 in a corresponding non-diseased sample.

Statistical methods for determining appropriate control or threshold values will be readily apparent to those of ordinary skill in the art. The control or threshold values may have been determined, if necessary, from samples of sub jects of the same age group, demographic features, and/or disease status, etc. The negative control or threshold value may originate from a single individual not af fected by a cancer in question or a subtype thereof, or be a value pooled from more than one such individual.

As used herein, the term "increased expression" refers to an increase in the amount of a biomarker in a sample as compared with a relevant control. Said increase can be determined qualitatively and/or quantitatively according to standard methods known in the art. The expression is increased if the amount or level of the biomarker in the sample is, for instance, at least about 1.5 times, 1.75 times, 2 times, 3 times, 4 times, 5 times, 6 times, 8 times, 9 times, time times, 10 times, 20 times or 30 times the control value or the amount of the same bi omarker in the control sample ln some embodiments, the term "increased ex pression" refers to a statistically significant increase in the level or amount of the biomarker as compared with that of a relevant, preferably negative, control. In creased expression as compared with a negative control indicates that the subject has or is at increased risk of having poor prognosis or at increased risk of relapse.

As used herein, the term "non-increased expression" refers to an ex pression level of a biomarker that is essentially the same both in a sample to be analyzed and in a relevant control ln some embodiments, non-increased expres sion as compared to a negative control indicates that the subject does not have or is not at risk of having poor prognosis or is not at risk of relapse. On the other hand, corresponding expression level as compared to a positive control indicates that the subject is at risk of having poor prognosis or is at risk of disease relapse lf compared to a healthy or apparently healthy control, non-increased expression may be referred to as normal or unchanged expression. As used herein, the term "decreased expression" refers to a decrease in the amount of a biomarker in a sample as compared with a relevant control. Said decrease can be determined qualitatively and/or quantitatively according to standard methods known in the art. The expression is decreased if the amount or level of the biomarker in the sample is, for instance, at least about 1.5 times, 1.75 times, 2 times, 3 times, 4 times, 5 times, 6 times, 8 times, 9 times, time times, 10 times, 20 times or 30 times lower that the control value or the amount of the same biomarker in the control sample ln some embodiments, the term "de creased expression" refers to a statistically significant decrease in the level or amount of the biomarker as compared with that of a relevant control. Decreased expression as compared with a positive control may indicate that the subject does not have or is not at risk of having poor prognosis.

As used herein, the term "sample" refers to a biological or clinical sample obtained from a subject whose ARPP19 expression level is to be deter mined. Preferred sample types include bone marrow samples obtained e.g. by a biopsy or aspiration, and blood samples, such as whole blood, serum, plasma, fractionated or non-fractionated peripheral blood mononuclear cells (PBMCs) or any purified blood cell type.

The term "sample" also includes samples that have been manipulated or treated in any appropriate way after their procurement, including but not lim ited to centrifugation, filtration, precipitation, dialysis, chromatography, treat ment with reagents, washing, or enriching for a certain component of the sample such as a cell population.

As used herein, the term "subject" refers to any mammal including, but not limited to, humans and domestic animals such as livestock, pets and sporting animals. Examples of such animals include without limitation carnivores such as cats and dogs and ungulates such as horses. Thus, the present invention may be applied in both human and veterinary medicine. As used herein, the terms "sub ject", "patient" and "individual" are interchangeable.

As used herein, the term "apparently healthy" refers to an individual or a pool of individuals who show no signs of haematological cancer such as leu kaemia, especially AML, and thus are believed not to be affected by said haemato logical cancer and/or who are predicted not to develop said haematological can cer.

As used herein, the term "haematological cancer" refers to a malignan cy affecting blood, bone marrow or lymph nodes. Haematological cancers are re- ferred to as leukaemia, lymphoma and myeloma depending on the type of cell af fected. Lymphoma is a group of blood cancers that develop from lymphocytes. Multiple myeloma, also known as plasma cell myeloma, is a cancer of plasma cells.

As used herein, the term "leukaemia" refers to a cancer which starts in blood-forming tissue, usually the bone marrow, and affects white blood cells. Leukaemia can be classified by the type of white cell affected (myeloid or lym phatic). Non-limiting examples of leukaemias include acute myeloid leukaemia (AML), chronic myeloid leukaemia (CML), acute promyelocytic leukaemia (APL), and subgroups thereof.

As used herein, the term "minimal residual disease" (MRD) refers to a low-level disease, wherein small numbers of leukaemic cells, usually below the detection limit, remain during treatment or after treatment when the patient is in remission (no symptoms or signs of disease). MRD is the major cause of relapse in leukaemia.

As used herein, the term "relapse" refers to a recurrence of a past med ical condition or the signs and symptoms thereof after a period of improvement or remission.

As used herein, the term "remission" refers to the disappearance of the signs and symptoms of a disease. A remission can be temporary or permanent.

As used herein, the term "prognosis" refers to a probable course or clinical outcome of a disease, while the expressions "prognosticating" and "prog nosing" refer to a prediction of future course of the disease.

As used herein, the term "good prognosis" refers to a probable favour able course of the disease for a certain period of time. "Prolonged overall surviv al", "prolonged disease-free survival", "prolonged recurrence-free survival" and "prolonged progression-free survival", when compared to the median outcome of the disease for example, are non-limiting examples of good prognosis. Depending on the haematological cancer and embodiment in question, good prognosis may refer to, for example at least 20%, at least 30%, at least 40%, at least 50%, at least 60%, at least 70%, at least 80% or at least 90% chance of surviving more than 1 year, more than 2 years, more than 3 years, more than 4 years or more than 5 years after the diagnosis.

As used herein, term "poor prognosis" refers to a probable non- favourable course of the disease for a certain period of time. "Reduced overall survival", "reduced disease-free survival", "reduced recurrence-free survival" and "reduced progression-free survival", when compared to the median outcome of the disease for example, are non-limiting examples of poor prognosis. Depending on the haematological cancer and embodiment in question, poor prognosis may refer to, for example less than 60%, less than 50%, less than 40%, less than 30%, less than 20% or less than 10% a likelihood of surviving more than 1 year, more than 2 years, more than 3 years, more than 4 years or more than 5 years after the diagnosis.

Likelihood of a certain clinical outcome of a disease, such as likelihood of a poor prognosis in a subject diagnosed with haematological cancer, may be expressed as a probability, such as a "hazard ratio" (HR), which term refers to the ratio of an instantaneous risk in two experimental arms over the study time peri od. lt represents point estimate at any given point of time.

Thus, with respect to poor prognosis, for example, if the HR risk esti mate for a particular biomarker level is greater than one, this indicates that a sub ject with this particular biomarker level has a higher risk for poor prognosis than a subject without the particular biomarker level ln contrast, if the HR risk esti mate for a particular biomarker level is less than one, this indicates that a subject with this particular biomarker level has a reduced risk for the poor prognosis as compared to a subject without the particular biomarker level.

As used herein, the term "95% confidence interval" (95% Cl) refers to an estimated range of values that one can be 95% certain contains the true mean of the population, the estimated range being calculated from a given set of sample data. Generally, the 95% Cl is used to estimate the precision of the HR. A wide Cl indicates a low level of precision of the HR, whereas a narrow Cl indicates a high er precision of the HR.

Univariate and multivariate analyses described herein revealed statis tically significant associations between ARPP19 expression levels and clinical outcomes, such that patients with lower ARPP19 burden survived longer and had longer relapse-free survival times ln an embodiment described in the experiment part, five-year relapse rate was 7% for the patients in the group of lowest quartile of ARPP19 expression while five-year relapse rate was 33% for patients that had ARPP19 expression higher than the lowest quartile. To put it differently, five-year relapse rate was 17% for the patients that had lower than median ARPP19 ex pression while five-year relapse rate was 35% for patients that had ARPP19 ex pression higher than median lmportantly, ARPP19 was a significantly stronger predictor of relapse-free survival than EV11 in a univariate analysis and especially in a multivariate analysis. Moreover, ARPP19 was a significantly stronger predic- tor of overall survival than EV11 in a univariate analysis. Results of the univariate and multivariate analyses are shown in Tables 7 and 8 below.

ln summary, increased expression level of ARPP19 is an indicator of poor prognosis, such as reduced overall survival or reduced relapse-free survival; whereas non-increased or decreased expression level of ARPP19 is an indicator of good prognosis, such as increased overall survival or increased relapse-free sur vival.

Turning now to Figure 2, it demonstrates that the AML patient cohort used in the studies leading to the present invention is a representative and accu rate patient cohort because a risk stratification analysis of the cohort according to the European LeukemiaNet recommendation (see, Table 2) could correctly strati fy the patents into favourable, intermediate and adverse risk groups.

lmportantly, it was found out that increased expression of ARPP19 is an adverse prognostic marker of haematological cancers, especially leukaemias, such as AML, independent from the classical risk stratification according to the European LeukemiaNet recommendation (p=0.83 by restricted maximum likeli hood (REML) estimation of variance). On the other hand, patients with lowest quartile expression of ARPP-19 have significantly lower risk for disease relapse over 12 years follow-up period (Fig. 4B).

Therefore, ARPP19 determinations may provide substantial help in clinical decision making in choosing appropriate treatment procedures, especially for subjects belonging to the intermediate risk group. For example, increased ex pression of ARPP19 in a subject belonging to the intermediate AML risk group suggests that it might be beneficial to direct such subjects to immediate stem cell transplantation, which is a treatment modality well known in the art. On the other hand, it would be sufficient to treat AML patients belonging to the intermediate risk group and having non-increased or downregulated expression of ARPP19 with chemotherapy only. Thus, ARPP19 may be used for stratifying subjects into groups of either poor prognosis or good prognosis, as well as stratifying subjects for different treatment modalities.

ln one aspect, the invention provides an in vitro method of selecting suitable treatment regime to a subject diagnosed with AML, the method compris ing the steps of: determining the expression level of ARPP19 in a biological sam ple obtained from the subject; determining whether or not the determined ex pression level of ARPP19 is increased in the sample by comparing said expression level to a control; and selecting a treatment regime comprising stem cell therapy if said expression is increased, or selecting a treatment regime comprising chemo therapy if said expression is non-increased (i.e. normal or unchanged) or de creased. Notably, said method does not involve therapeutic intervention but only provides help in clinical decision-making. Typically, said control is the level of ARPP19 expression in an apparently healthy subject.

ln some other implementations, the present method of determining the probable outcome of haematological cancer may further include therapeutic intervention. Once an individual is identified to have a given probable outcome of the disease, he/she may be subjected to an appropriate therapeutic intervention, such as stem cell therapy and/or chemotherapy. Thus, in some embodiments, the present invention provides a method of treating haematological cancer, especially leukaemia such as AML, in a subject in need thereof, wherein a biological sample obtained from the subject is contacted with a reagent that specifically detects ARPP19, and determining whether or not the expression of ARPP19 is increased, non-increased or decreased in said sample by comparing said expression with a relevant control lf said expression is increased, the subject should be assigned to or treated with a treatment regime comprising stem cell therapy, whereas if said expression is non-increased or decreased, the subject should be assigned to or treated with a treatment regime comprising chemotherapy. To put is differently, the method comprises administering stem cell therapy to subjects with increased ARPP19 expression as compared to a control, while administering chemotherapy to subjects with non-increased (i.e. normal) or decreased ARPP19 expression.

ln some embodiments, the prognosis of haematological cancer on the basis of ARPP19 expression is not finally determined ln such embodiments, the method is not by itself determinative of the prognosis of cancer, but can indicate that further prognostic testing is needed or would be beneficial. Therefore, ARPP19 may be used in combination with one or more other prognostic methods or markers for the final determination of the prognosis. As readily understood by those skilled in the art, this applies to other aspects of the invention as well, such as predicting relapse or determining risk of relapse.

Turning now to Figure 1, it demonstrates that the expression levels of known AML-associated biomarkers WT1 and EV11 in the AML patient cohort used in the studies leading to the present invention are in concert with published ex pression levels of these biomarkers in AML patients (Hinai & Valk, 2016, ibid.) Ujj et al., 2016, ibid.).

Accordingly, in some embodiments of the invention, the measurement of ARPP19 can be performed in combination with (i.e. prior to, simultaneously or after) determining the expression level of other markers relevant to the given haematological cancer. Such methods and markers are well known to a person skilled in the art. With respect to AML, such other markers include, but are not limited to, WT1 and EV11.

Receiver Operation Characteristic (ROC) curves may be utilized to demonstrate the trade-off between the sensitivity and specificity of a marker, as is well known to skilled persons. The horizontal X-axis of the ROC curve represents 1 -specificity, which increases with the rate of false positives. The vertical Y-axis of the curve represents sensitivity, which increases with the rate of true positives. Thus, for a particular cut-off (i.e. threshold) selected, the values of specificity and sensitivity may be determined ln other words, data points on the ROC curves represent the proportion of true-positive and false-positive classifications at vari ous decision boundaries. Optimum results are obtained as the true-positive pro portion approaches 1.0 and the false-positive proportion approaches 0.0. Howev er, as the cut-off is changed to increase specificity, sensitivity usually is reduced and vice versa.

The area under the ROC curve, often referred to as the AUC, is a meas ure of the utility of a marker in the correct identification of disease subjects, i.e. subjects who are affected by a disease. Thus, the AUC values can be used to de termine the effectiveness of the test. An area of 1.0 represents a perfect test, whereas an area of 0.5 represents a worthless test. However, an AUC value is not the only parameter to be used in determining the potential of a biomarker for clinical use. ln some cases, a biomarker may be clinically valuable even if it has a relatively low AUC value.

ln the analyses described herein, determination of ARPP19 levels alone provided an AUC value of 0.667 for relapse-free survival, while the same obtained with EV11 was 0.686. On the other hand, when ARPP19 and EV11 were used in combination, the AUC value increased to 0.771. Thus, in some embodi ments, it may indeed be beneficial to use ARPP19 in combination with one or more other markers, such as EV11, as already set forth above.

ln some embodiments, the prognosis may be based on analyzing one or more serial samples obtained from the subject, for example, to detect any changes in the prognosis, and may involve a prediction of a response to a particu lar treatment or combination of treatments for haematological cancer ln such in stances, the prognostication method comprises analyzing and comparing at least two samples obtained from the same subject at various time points. The number and interval of the serial samples may vary as desired. The difference between the obtained assessment results serves as an indicator of the course of the haemato- logical cancer and risk of relapse. ARPP19 may thus be used for various monitor ing purposes such as monitoring for cancer progression over time, monitoring for or determining any possible cancer remission, monitoring for or determining any possible recurrence or relapse, monitoring for or determining the presence of minimal residual disease, e.g. during the first remission phase after intense chem otherapy and monitoring for or determining response to treatment. Further uses of ARPP19 include, but are not limited to, patient grouping or stratification.

Accordingly, the present invention provides an in vitro method of pre dicting relapse in a subject having haematological cancer, such as leukaemia, es pecially ALM, wherein the method comprises: assaying the level of ARPP19 ex pression in samples obtained from the subject at two or more time points; detect ing a change (if any) in the assayed expression levels of ARPP19 between the two or more time points; and predicting the relapse from the change in the levels of the ARPP19 expression, wherein increased risk of relapse is determined from an increase in the level of ARPP19 expression. On the other hand, non-increased risk of relapse is determined from a non-increased level of ARPP19 expression ln some embodiments, the samples are obtained from the subject after remission and used for monitoring for any change in the prediction of the relapse risk ln some embodiments, the samples to be assayed and used in the comparisons to detect any change in the level of ARPP19 expression may include one or more samples obtained at diagnosis and/or during treatment.

The above-disclosed method of predicting relapse may also be formu lated as a method of determining relapse risk, as readily understood by those skilled in the art.

Also provided is an in vitro method of determining the risk of relapse in patients with MRD in a subject having haematological cancer, such as leukae mia, especially AML, wherein the method comprises: assaying the level of ARPP19 expression in samples obtained from the subject at two or more time points; de tecting a change (if any) in the assayed expression levels of ARPP19 between the two or more time points; and determining the risk of relapse from the change in the levels of the ARPP19 expression, wherein the increased risk for forthcoming relapse is determined from an increase in the level of ARPP19 expression as com pared to earlier measurements from the same patient having MRD. On the other hand, the non-elevated risk of relapse is determined from a non-increased level of ARPP19 expression as compared to earlier measurements from the same patient having MRD. ln some embodiments, the samples are obtained from the subject after remission and used for determining risk of forthcoming relapse ln some embodiments, the samples to be assayed and used in the comparisons to detect any change in the level of ARPP19 expression may include one or more samples obtained at diagnosis and/or during treatment.

As readily understood by those skilled in the art, also the other uses of ARPP19 mentioned above may be formulated as methods for the indicated pur poses.

ln the methods disclosed above, the step of assaying the level of ARPP19 expression in a one or more samples obtained from the subject may be carried out by contacting the sample with a reagent specific for ARPP19 and thus suitable for determining the expression level of ARPP19. Non-limiting examples of suitable reagents and techniques for determining ARPP19 expression are dis closed below.

As used herein, the expression "a change in the assayed expression levels of ARPP19 between the two or more time points" refers to a comparison of the expression level in an earlier sample to the expression level in a sample ob tained at a later time point. Thus, in this context, the term "increased" expression means that the expression level in the later sample is higher than in the earlier sample.

The present invention provides advantages not only for individual pa tient care but also for better selection and stratification of patients for clinical tri als. For example, cancer patients grouped on the basis of their ARPP19 expression levels could be employed in clinical studies with the aim of developing efficient new personalized therapeutic tools such as new medical procedures or drugs.

For example, predicting response to treatment or determining any change in the prognosis may be carried out in any of the following cohorts: a co hort that responds favourably to a treatment regimen, a cohort that does not re spond favourable to a treatment regimen, a cohort that responds significantly to a treatment regimen, a cohort that does not respond significantly to a treatment regimen, and a cohort that responds adversely to a treatment regimen (e.g. exhib its one or more side effects). These cohorts are provided as examples only, and any other cohorts or sub-cohorts may be analysed. Thus, a subject may be deter mined for ARPP19 expression level to predict whether he/she will respond fa- vourably or significantly to a treatment regiment, not respond favourably or sig nificantly to a treatment regimen, or respond adversely to a treatment regimen lf a subject is likely to respond positively to a cancer treatment, said subject may be treated accordingly. On the other hand, if a subject is likely to respond negatively (i.e. not exhibit positive effects or exhibit adverse side effects) to a cancer treat ment, an alternative course of treatment may be applied.

Determining the expression level of ARPP19 in a sample obtained from a subject whose haematological cancer is to be prognosed or who is to be moni tored e.g. for cancer progression, cancer remission, recurrence, relapse, response to treatment, or who is to be subjected to therapy selection or grouped or strati fied for any relevant cancer-related purpose, can be determined by any available or future means suitable for this purpose. Said determination may be executed at different molecular levels, preferably at mRNA, cDNA, or protein level as is well known in the art. The present invention is not limited to any determination tech nique. Thus, in different embodiments, different means or combinations thereof for analysing a biological or clinical sample for the expression of ARPP19 may be used.

Generally, determining the level of ARPP19 expression at protein level comprises contacting a sample obtained from a subject in need of said determina tion with a binding body, such as an antibody, specifically recognizing ARPP19 polypeptide under conditions wherein the binding body specifically interacts with ARPP19; and detecting said interaction (if any); wherein the presence or degree of said interaction correlates with the presence of ARPP19 or the level of ARPP19 expression in said sample. Binding bodies alternative to antibodies and fragments thereof suitable for determining ARPP19 expression at protein level include but are not limited to oligonucleotide or peptide aptamers, receptors and biologically interacting proteins.

Accordingly, in some embodiments the present invention, expression level of ARPP19 may be determined by an immunoassay, which comprises for ex ample the following steps: providing an antibody that specifically binds to ARPP19; contacting a patient sample, preferably comprising permeabilied or dis rupted cells, with anti-ARPP19 antibody; and detecting the presence of a complex of the antibody bound to ARPP19 (if any) in the sample lf desired, the antibody can be fixed to a solid support to facilitate washing and subsequent detection of the complex, prior to contacting the antibody with the sample. After incubating the sample with an anti-ARPP19 antibody, the mixture may be washed and the antibody-ARPP19 complex formed can be detected. This can be accomplished by e.g. using a detectable antibody, i.e., detectably labeled antibody, or an antibody labelled with an enzyme and incubating the complex with a detection reagent, i.e. substrate of the enzyme. Alternatively, ARPP19 can be detected using an indirect assay, wherein, for example, a second, labeled antibody is used to detect the anti- body-ARPP19 complex formed.

ln some embodiments, the expression level of ARPP19 may be deter mined using a non-competitive assay format. Non-competitive immunoassays, also known as reagent excess assays, sandwich assays, immunometric assays or two-site assays, generally involve use of two antibodies targeting different epitopes in the antigen, one antibody for antigen capture and the other labeled for detection. A person skilled in the art will be well capable of establishing a binding assay for measuring the level of ARPP19 alone or in relation with any control. Ac cordingly, formation of the antibody-ARPP19 complex may be determined using any of a number of well-recognized non-competitive immunological binding as says. Useful assays include, for example, enzyme immune assays (EIA) such as en zyme-linked immunosorbent assay (ELISA); fluorescent immunosorbent assays (FLA) such as time-resolved immunoflurometric assays (TR-1FMA); chemilumi nescence immunoassays (CL1A) and radioimmune assays (R1A).

Depending on the assay type employed, either the first antibody or the second antibody, or both, may be conjugated or otherwise associated with a de tectable label selected from the group including, but not limited to, optical agents such as fluorescent labels including a variety of organic and/or inorganic small molecules and a variety of fluorescent proteins and derivatives thereof, phospho rescent labels, chemiluminescent labels, and chromogenic labels; radioactive la bels such as radionuclides that emit gamma rays, positrons, beta or alpha parti cles, or X-rays, and enzymes such as alkaline phosphatase (AP) and (horseradish) hydrogen peroxidase (HRP). Said association can be direct, e.g. through a covalent bond, or indirect, e.g. via a secondary binding agent, a chelator, or a linker. Tech niques for conjugating or otherwise associating detectable agents to antibodies are well known and antibody labelling kits are commercially avail-able from doz ens of sources. One or both of the antibodies may also be expressed as fusion pro teins with a detectable label or a detection tag by recombinant techniques.

ln some embodiments of non-competitive immunoassays, the detec tion antibody is detectably labelled ln some other embodiments, the detection antibody is recognized by a further antibody comprising a detectable label ln some still other embodiments, the detection antibody comprises a tag that is rec ognizable by a further antibody comprising a detectable label ln some still fur ther embodiments, the detection antibody and said further antibodies are labelled with the same label, e.g. for improving sensitivity ln some other embodiments, the detection antibody and said further antibodies are labeled with different la bels.

lmmunoassays suitable for carrying out the method of the present in vention include solid-phase immunoassays, such as lateral flow assays and con ventional sandwich assays carried out on a solid surface such as glass, plastic, ce ramic, metal or a fibrous or porous material such as paper, in the form of e.g. a microtiter plate, a stick, a card, an array, a sensor, a bead, or a microbead. Said solid-phase immunoassay may be either heterogeneous or homogeneous ln het erogeneous assays, any free antigens or antibodies are physically separated from immunocomplexes formed, e.g. by washings, while no such separation is neces sary in homogeneous assays including the many forms of biosensors.

Suitable homogeneous immunoassays are not limited to solid-phase assay formats but encompass also homogeneous immunoassays carried out in so lution. Such in-solution immunoassays are particularly advantageous because nei ther immobilization nor washing steps are required, making them simple and easy to perform. Thus, in some embodiments, the immunoassay is liquid-based homogeneous immunoassay.

Further examples of suitable methods for determining the expression level of ARPP19 at protein level include conventional Western blot assays, dot blot assays and assay formats based on immunohistochemistry. Moreover, for ex ample flow cytometry, such as fluorescence-activated cell sorting (FACS), may be used for detecting antibody-ARPP19 complexes. Moreover, techniques suitable for determining the expression level of ARPP19 at protein level include MS-based detection methods, such as selected reaction monitoring (SRM), SWATH and mass cytometry from cancer cell samples based on the unique peptide. Also in such embodiments, the method may include comparing the expression level of ARPP19 with that of a control biomarker.

Applicable methods for use in determining ARPP19 expression at nu cleic acid level include but are not limited to microarrays, which are a collection of nucleic acid, e.g. DNA, spots attached to a solid surface. Each DNA spot contains an oligonucleotide of a specific DNA sequence, known as a probe or capture probe. These can consist of a short section of a gene or other oligonucleotide such as DNA or LNA element that are used to hybridize target cDNA, mRNA or cRNA (also called anti-sense RNA) in a sample under high-stringency conditions. Probe- target hybridization can be detected and/or quantified e.g. by detection of fluoro- phore-, silver-, or chemiluminescence-labeled target nucleic acids to determine their relative abundance in the sample ln some array types a recognition element (transducer) capable of converting a surface-bound probe-target nucleic acid in teraction into a quantifiable signal may be incorporated. Among the several dif ferent types of transducers available, especially those based on electrochemical or optical detection are popular in the art. Similarly to other nucleic acid analysis methods, both types of sensors can either be run as label-free or label-using and also divided to heterogeneous and homogeneous assay formats based on the re quirement for washing.

One suitable application is the NanoString's nCounter technology, which is a variation on the DNA microarray lt uses molecular "barcodes" and mi croscopic imaging to detect and count up to several hundred unique transcripts in one hybridization reaction. Each color-coded barcode is attached to a single tar- get-specific probe corresponding to a gene of interest ln some embodiments the use of an array based technology such as the NanoString is beneficial in order to detect ARPP19 expression.

ln accordance with the above, the expression of ARPP19 at nucleic acid level may be determined using any suitable method with or without nucleic acid amplification. For example, ARPP19 mRNA may be first converted into its com plementary cDNA with the aid of a reverse transcriptase, followed by DNA ampli fication, e.g. by reverse transcriptase PCR (RT-PCR) including but not limited to quantitative PCR (qPCR), also known as real-time PCR. The presence, absence or concentration of the expressed ARPP19 mRNA polynucleotide or an amplification product thereof may be assessed according to methods available in the art, for example by using a ARPP19-specific capture or detection probe. Further potential methods suitable for determining the expression of ARPP19 at nucleic acid level, with or without a reverse transcription step and/or a nucleic acid amplification step depending on the method selected, include but are not limited to RNase pro tection assays, molecular beacon-based oligonucleotide hybridization assays, melting curve analysis combined with oligonucleotide probes and/or intercalat ing labels, gel electrophoresis analysis, Southern blotting, microarrays such as DNA microarrays and RNA microarrays, direct probing, and signal accumulation assays. The detection of hybridization complexes can be carried out by any suitable method available in the art. Hybridization complexes may in some assay methods be physically separated from unhybridized nucleic acids (e.g. by employ ing one or more wash steps to wash away excess target mRNA/cDNA, the probe, or both), and detectable labels bound to the complexes are then detected. More often, however, homogeneous detection is employed where no physical separa tion of hybridized and unhybridized molecules is required ln nucleic acid ampli fication assays, homogeneous detection may be performed as a separate step af ter amplification, i.e. using homogeneous end-point detection. The accumulation of the amplicon can also be homogeneously monitored during the amplification assay by use of qPCR type of methods. The various homogeneous detection meth ods can further be classified into probe-using and non-probe-using formats.

Detectable labels refer to radioactive, fluorescent, biological or enzy matic tags or labels of standard use in the art. A detectable label can be conjugat ed to either the oligonucleotide probe or the target polynucleotide. As is evident to those skilled in the art, the choice of a particular detectable label dictates the manner in which it is bound to the probe or the target sequence, as well as the technique to be used for detection. Although the present invention is not specifi cally dependent on the use of a label for the detection of a particular nucleic acid sequence, such a label might be beneficial, by increasing the sensitivity or speci ficity of the detection and simplifying the multiplexing of the assays. Furthermore, a detectable label may better enable automation.

lntercalating dyes (e.g. SYBR Green) can be used to detect the amplifi cation of the DNA fragment of interest during a nucleic acid amplification reaction such as PCR. lntercalation occurs when ligands of an appropriate size and chemi cal nature position between the planar base pairs of DNA. These ligands are most ly polycyclic, aromatic, and planar, and therefore often make suitable nucleic acid stains. The intensity of fluorescence increases respectively during the amplifica tion and it can be measured in real-time without the need of separate oligonucleo tide probes.

As appreciated by those skilled in the art, a variety of controls and ad ditives may be employed to improve accuracy of hybridization assays. For in stance, samples may be hybridized to an irrelevant probe and/or treated with RNAse A prior to hybridization, to assess false hybridization or prevent unspecific or unwanted effects.

ln some embodiments, determining the expression of ARPP19 is per- formed by sequencing techniques. Numerous methods suitable for this purpose have been described in the art and include, but are not limited to, traditional Sanger sequencing and next-generation sequencing (NGS) techniques. The pre sent embodiments are not limited to any branded technique.

A representative commercial platform suitable for use in accordance with some embodiments of the invention, wherein the presence or quantity of ARPP19, if any, is detected by sequencing, is lllumina’s sequencing by synthesis (SBS) technology, particularly TruSeq® technology. Applying TruSeq® technolo gy requires that two oligonucleotide probes, which hybridize upstream and downstream of the region of interest, are designed and synthetized. Each probe contains a unique, target specific sequence and a universal adapter sequence. An extension-ligation reaction is used to unite the two probes and create a library of new template molecules with common ends. Adapter-ligated DNA is then subject ed to PCR amplification, which adds indexes and sequencing primers to both ends. Sequencing may then be performed by any suitable equipment, such as MiS- eq® sequencer, utilizing a reversible terminator-based method enabling detec tion of single bases as they are incorporated into growing DNA strands.

Non-limiting examples of suitable equipment for the present sequenc ing purposes include lllumina® Sequencers, such as MiSeq™, NExtSeq500™, and HiSeq™ (e.g. HiSeq™ 2000 and HiSeq™ 3000), and Life Technologies’ Sequencers, such as lon Torrent™ Sequencer and lon Proton™ Sequencer lt should be under stood that utilizing any of these equipment requires that appropriate sequencing technique and chemistry be used.

ln some embodiments relating to NGS techniques, detection of ARPP19 mRNA is possible with deep sequencing such as a one performed with lllumina HiSeq™ 3000 platform, 150 bp reading length and paired-end library.

Further methods suitable for detecting the expression level of ARPP19 include, but are not limited to, RNA in situ hybridization technologies such as RNAscope® (Advanced Cell Diagnostics, ACD) and ViewRNA™ (lnvitrogen).

ln a further aspect, the present invention provides a kit and use there of for detecting the expression level of ARPP19 in a biological or clinical sample, for any purpose set forth above. Preferably, the kit comprises a ARPP19-specific binding body, such as an anti-ARPP19 antibody or at least one ARPP19-specific oligonucleotide, such a probe and/or a primer pair ln some embodiments, the binding body may comprise a detectable label. A person skilled in the art can easi ly determine any further reagents to be included in the kit depending on the de- sired technique for carrying out determination of the expression level of ARPP19. Thus, the kit may further comprise at least one reagent for performing for exam ple an immunoassay such as EL1SA, a Western blot, immunohistochemical assay, nucleic acid amplification assay, signal amplification assay, in situ RNA hybridiza- tion assay, mass spectrometric assay or sequencing assay.

ln some embodiments, an appropriate control reagent or sample or a threshold value may be comprised in the kit. The kit may also comprise a com puter readable medium, comprising computer-executable instructions for per forming any of the methods of the present disclosure.

lt will be obvious to a person skilled in the art that, as technology ad vances, the inventive concept can be implemented in various ways. The invention and its embodiments are not limited to the examples described below but may vary within the scope of the claims.

EXPERIMENTAL PART

Materials and Methods

Patient cohort

Patient cohortl: Patient samples for this study were collected between January 2000 and July 2010, a total of 80 patients aged 18-65 (median 50 years (Qi=38.8, Q₃=58.0)) diagnosed with de novo or secondary AML at Turku Universi- ty Central Hospital (TYKS). Patients with t(15;17) (q22;ql2) were excluded from this study. Patient characteristics are presented in Table 1. Median overall surviv al in the whole cohort was 5.4 years (95% Cl 2.8-7.9). Karyotyping and identifica tion of prognostic subgroups within AML was conducted by TYKSLAB, which is a member of the European Leukemia Net (ELN). The applied risk stratification based on cytogenetic and molecular findings is in close accordance with the Eu ropean LeukemiaNet recommendation (Table 2). Most patients (76) were en rolled in the Finnish Leukemia Study Group prospective protocols (Table 3). 32 patients were treated according to AML92 and 44 according to AML2003 proto col. Treatment of 4 patients was significantly modified due to patient related rea- sons. Although patients were treated with different schedules, all received regi mens based on anthracycline and cytarabine as induction therapy. High-dose cy- tarabine and allogenic stem cell transplantation when possible, were used as con solidation therapy. No significant difference was found related to relapse or OS rates of patients with AML92 or AML2003 treatment protocols lnformed consent was obtained from all patients and the local Ethical Review Board of TYKS ap proved the study protocol. Comprehensive clinical data with median 5.4 years (range 6 days - 16 years) of follow-up has been gathered from the patients. No missing data imputation was required since the analyzed dataset was complete. Table 1. Clinical and molecular characteristics of a series of 80 patients with AML

Table 2. Risk stratification that was applied in retrospective analyses (modified from LeukemiaNet; Dohner et al. 2010)

Table 3. Chemotherapy according to AML92 and AML2003 protocols

Patient cohort2: 123 AML bone marrow patient samples were ana lyzed from Finnish Hematology Registry and Biobank (FHRB) collection. Only 48 patients had received intensive chemotherapy as an induction therapy and from which complete clinical data was available. Patient characteristics are presented in Table 4. Samples were collected from Finnish university hospitals and other hematological units between December 2011 and April 2016. Median age for the 48 patients was 58.2 years (Qi=48.7, Q₃=67.3), median overall survival was 2.5 years (95% Cl 1.5-3.4) and median follow-up time was 2.5 years (range 7 days - 5.2 years). FHRB is authorized by the National Supervisory Authority for Welfare and Health (Valvira) and has got a favorable opinion from the National Medical Ethics Committee.

Table 4. Clinical and molecular characteristics of a series of 48 patients with AML from FHRB

RNA isolation and cDNA synthesis

Total RNA was isolated from extracted mononuclear cells (patient bone marrow samples). Total RNA was extracted using the E.Z.N.A.® Total RNA Kit 1 (Omega Bio-Tek Inc, Norcross, GA, USA) according to the manufacturer’s in structions. After isolation, RNA concentration was measured using a NanoDrop spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). cDNA synthe sis for all samples was performed using 1 gg of total RNA as a starting material. cDNA of patient samples was synthesized using Superscript 111 Reverse Tran scriptase (18080093, lnvitrogen, Carlsbad, CA, USA), random primers (C1181, Promega), RiboLock(tm) Ribonuclease lnhibitor (#E00381, Thermo scientific) and dNTP-mix (B10-39028, Bioline, London, UK). RT-reactions, including temper atures and volumes, were performed according to enzyme’s manufacturer's in structions.

Quantitative real-time PCR

Primers for each gene specific assay, if possible, were designed to be located to different exonic sequences to avoid amplification of genomic DNA. Pri mer concentration in each reaction was 300 nM and probe concentration 200nM. Specificity of qPCR reactions was verified by agarose gel electrophoresis and melting curve analysis. One band of the expected size and a single peak, respec tively, were required. The efficiency of amplification for each target was also as sessed with standard curve analysis. Amplification of target cDNAs was per formed using KAPA PROBE FAST qPCR Kit (Kapa Biosystems) and 7900 HT Fast Real-Time PCR Sys-tem (Thermo Fisher) according to the manufacturers’ instruc tions. Quantitative real-time PCR was executed under the following conditions: 95°C for 10 min followed by 45 cycles of 95°C for 15 s and 60°C for lmin. Relative gene expression data was normalized to expression level of endogenous house keeping genes Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and beta acting using 2^A-AAC(t) method with SDS software (version 2.4.1, Applied Biosys tems) or with Thermo Fisher Cloud (Patient cohort2). To estimate degree of over expression in AML, expression of each gene was normalized to expression level in commercial normal, pooled bone marrow control sample from 56 males and fe- males (636591, lot 1002008, Clontech). Results were derived from the average of at least two independent experiments and two technical replicates.

Primer and probe sequences used for qPCR analysis are listed in Table 5 below. For SET detection, custom TaqMan® gene expression assay A1Y9AXG (Thermo Fisher Scientific) was used.

Table 5. Primer and probe sequences used

* Roche Universal ProbeLibrary (UPL) probe

Statistical analysis

Statistical analysis and comparisons were performed using SAS soft- ware (version 9.3 for Windows, SAS lnstitute lnc., Cary, NC, USA) or JMP pro (ver- sion 12.0 for Windows software, SAS lnstitute Inc.). All statistical tests were two- sided and declared significant at a p-value of <0.05. Continuous variables were summarized by descriptive statistics (median, interquartile range and range) while frequencies and percentages were calculated for categorical data. For con- tinuous variables, when possible, transformations were performed to achieve a normal distribution assumption. Wilcoxon rank sum test and Student’s t-test were used for analyzing continuous variables. Frequency tables were analyzed using Fisher's exact test for categorical variables. Univariate survival analysis for overall survival (OS) and relapse-free survival (RFS) was based on the Kaplan- Meier method where stratum-specific outcomes were compared using log-rank (Mantel-Cox) statistics. To adjust for the covariates, a Cox proportional hazards regression model was used for univariate and multivariate analysis ln multivari ate analysis, covariates were entered in a stepwise backward manner.

OS was defined for all patients measured from the date of diagnosis to the date of death from any cause. Patients not known to have died at last follow up were censored on the date they were last known to be alive. RFS was defined for patients from the date of diagnosis until the date of relapse. Patients not known to have relapsed were censored on the date they were last examined.

Results Characterization of the patient cohort for ARPP19, WT1 and EVI1 mRNA ex pression

Quantitative real time PCR (qPCR) was used to analyze mRNA expres sion levels of ARPP19 in 80 bone marrow samples obtained from diagnosis phase AML patients. Patient characteristics are presented in Table 1 above. The samples were analyzed for expression of well-studied AML markers WT1 and EV11 mRNA expression as well. To estimate degree of overexpression in AML, expression of each gene was normalized to expression level in commercial normal, pooled bone marrow control sample from 56 males and females (Clontech). Expression level in the normal bone marrow sample was set as 1 and values lower than this were considered as lowexpression and values higher than this overexpression. Overex pression of measured genes as well as median expressions are shown in Table 6. Waterfall blots of the expression patterns of ARPP19, WT1 and EV11 are shown in Figure 1. Table 6. Overexpression of WT1, EVIland ARPP19 in 80 patients with AML

WT1 and EV11 showed expression patterns that are in accordance to published literature (Hinai & Valk, 2016, ibid.) Ujj et al., 2016, ibid.). WT1 mRNA expression was highly overexpressed in diagnosis phase AML patients bone mar row compared to normal bone marrow (Table 6 and Fig. 1A, 91%, n=73) as ex pected. ARPP19 overexpression in the sample panel was 21% (Fig. 1C, n=16) and EV11 overexpression was 13% (Fig. IB, n=10). Moreover, patients with ARPP19 overexpression showed higher incidence of NPM1 mutation (56% vs. 17%; p=0.003 by Fisher’s exact test) and were more likely to be classified as FAB Ml subtype (62% vs. 18%; p=0.009 by Fisher’s exact test). Patients that had both FLT3-1TD and NPM1 mutation (FLT3-1TD+, NPM1+) had higher ARPP19 gene ex pression than patients with FLT3-1TD but without NPM1 mutation (FLT3-1TD+, NPM1-; p=0.019 by Wilcoxon rank-sum test) or to patients without FLT3-1TD and without NPM1 mutation (FLT3-1TD-, NPM1-; p=0.021 by Wilcoxon rank-sum test). Although patients with ARPP19 overexpression had more likely normal karyotype (69% vs. 30%; p=0.008 by Fisher’s exact test) than patients with ARPP19 expression lower than normal control, ARPP19 expression levels were not significantly different between normal and other karyotype patients (Stu- dent’s t-test, p=0.23). ln our sample collection WT1 and EV11 gene expression were significantly different in the normal karyotype patients compared to other karyotypes (p=0.02 for both genes by Student’s t-test). Mean EV11 gene expres sion was lower and WT1 expression higher in normal karyotype patients.

Furthermore, diagnosis phase ARPP19 expression levels correlated with WT1 (r=0.42, p=0.001) gene expression. EV11 gene expression did not show any significant correlation with any other target gene in this patient cohort. Survival analysis based on EVI1 and ARPP19 mRNA expression in AML pa tients

Kaplan-Meier estimates were used to analyze the overall survival (OS) of the patients in cohort (Figure 2). Distribution of the patients to three clinically used risk groups (favorable n=21, intermediate n=37, adverse n=22) based on their genetic profiles is presented in Table 1. Five-year survival rate was 81% for the patients in favorable (Fig. 2, groupl), 51% for the patients in intermediate (group 2) and 27% for the patients in adverse risk group (group 3). The median OS was 5.4 years (95% Cl 2.8-7.9) and the probability of OS at five years was 52.5%. Risk group of the patients per se was a strong indicator of prognosis (Fig 2, p=0.003 by log-rank test), thus validating that our study material is highly rep resentative of standard clinical AML patient material. Additionally, high EV11 mRNA expression at diagnosis phase was a strong indicator of poor overall sur vival in this AML patient cohort (Fig. 3A, p=0.0004).

Cox univariate analysis revealed that EV11 expression at the diagnosis was the only qPCR marker in the panel that had significant role in predicting the OS of the patients (Table 7, p=0.013, HR 1.13 (95% Cl, 1.03 to 1.24)). Diagnosis age (p=0.0002, HR 1.08 (95% Cl, 1.04 to 1.12)) was also independent markers for OS prediction in this cohort. When dividing the cohort to patients that have died (n=39), or are alive (n=41), after the follow up time, only diagnosis age (57.5 years vs. 44.5 years, p<0.0001 by Wilcoxon rank-sum test) and diagnosis phase EV11 mRNA expression levels (0.0093 vs. 0.0023, p=0.037 by Wilcoxon rank-sum test) were significantly different between the two groups. Patients that died dur ing the follow-up time were older at diagnosis and had higher EV11 mRNA ex pression.

Additional analysis for the prognostic role of the studied genes for OS was performed by Cox’s proportional hazard model for OS, which included diag nosis age, FLT3-1TD status, NPM1 mutation status, and diagnosis phase mRNA expression levels of C1P2A, SET, EV11, WT1, ARPP19, T1PRL and PME1. ln the final model, diagnosis age (Table 8, p=0.0004, HR: 1.07), NPM1 mutation positivity (p=0.0165, HR: 0.21 (95% Cl, 0.057 to 0.75)) and EV11 expression (p=0.0263, HR: 1.14) were found as independent prognostic factor for OS. Notably, regarding PlPs, only ARPP19 mRNA expression was found as an independent prognostic factor for OS and its HR was found to be even higher than either EV11, or diagno sis age (p=0.0456, HR: 2.05 (95% Cl, 1.01 to 4.15)). Table 7. Univariate analysis for overall survival and relapse-free survival in entire patient cohort

Table 8. Multivariate analysis for overall survival and relapse-free survival in the entire patient cohort

High ARPP19 mRNA expression is an independent relapse marker

ln addition to overall survival markers, novel markers to predict pa tient relapse from intensive induction therapy would be in high demand in clini cal AML practice ln support to highly representative nature of our patient mate rial to identify novel relapse markers, the median follow-up time in 68 AML pa- tients who achieved CR was 7.2 years (range 0.2-15.9 years), and the patients that had relapse were more likely to be in adverse risk group than patients who did not have relapse during the follow-up time (40% vs. 14%, p=0.038 by Fisher’s exact test). There were no significant differences between non-relapsing and re lapsing groups when it comes to any other clinical characteristics, including the patient age, or either NPM1 mutation and FLT3-1TD gene fusion (Table 9). Table 9. Clinical and genetic characteristics of AML patients according to relapse

No relapse Relapse

Median ian

Characteristic N No. (%) p

(range) ge)

Overall 5 20 (28)

Sex

Male 2 11 (55) ns

Female 2 9 (45)

Age 49 (18-65) 9-63)

Diagnosis

AML-MO 2 (10) ns AML-M1 1 4 (20)

AML-M2 9 3 (15)

AML-M3 0 (0)

AML-M4 3 (15)

AML-M5 1 2 (10)

AML-M6 0 (0)

AML-M7 0 (0)

Not specified 6 6 (30)

Cytogenetic group (AML

2012)

Favorable 1 3 (15) 0.04

Intermediate 2 9 (45)

Adverse 7 8 (40)

Normal karyotype

No 2 15 (75) ns

Yes 2 5 (25)

Secondary AML

No 4 13 (65) ns

Yes 7 7 (35)

Extramedullary disease

No 4 18 (90) ns

Yes 7

2 (10) FLT3-ITD

No 40 (78) 18 (90) ns

Yes 9 (18) 1 (5)

Not analyzed 2 (4) 1 (5)

NPM1 mutated

No 32 (63) 17 (85) ns

Yes 17 (33) 2 (10)

Not analyzed 2 (4) 1 (5)

FLT3-ITD and NPM1 co

association

FLT3-ITD+, NPMlmutation+ 5 (10) 1 (5) ns FLT3-ITD+, NPMlmutation- 4 (8) 0 (0)

FLT3-ITD-, NPMlmutation+ 12 (23) 1 (5)

FLT3-ITD-, NPMlmutation- 28 (55) 17 (85)

Not analyzed 2 (3) 1 (0) ln regards to mRNA markers, EV11 (0.058 vs. 0.0026, p=0.023 by Wil- coxon rank-sum test), and ARPP19 (Figure 4A, 0.78 vs. 0.57, p=0.035 by Wilcoxon rank-sum test) diagnosis phase mRNA expression levels were significantly differ- ent between patients with relapse, and patient’s without relapse during the fol low-up time. There were no significant differences in the rate of CR (75% vs. 88%), resistance (25% vs. 8%) or death during induction therapy (0% vs. 3%) between patients with ARPP19 overexpression or underexpression.

Kaplan-Meier estimates were used to analyze the relapse-free survival (RFS) of the patients in cohort. As expected, risk group of the patients was a strong indicator of relapse (p=0.008 by log-rank test). Moreover, EV11 and ARPP19 gene expressions had a role in predicting the prevalence of relapse in this patient cohort. High EV11 mRNA expression was a strong indicator of shorter re- lapse-free survival (Fig. 3B, p<0.0001 by log-rank test). Furthermore, patients with the lowest quartile of ARPP19 expression were linked to longer RFS (Fig 4B, p=0.029; as compared to those over lowest quartile). Five-year relapse rate was 7% for patients with lowest quartile expression of ARPP19, while five-year re lapse rate was 33% for patients that had ARPP19 expression higher than the low est quartile. lnterestingly, patients in lowest quartile (Qi) ARPP19 expression represented all risk groups (Fig 4C), which underlines ARPP19’s role as a risk group independent marker for RFS. Additionally, Cox’s univariate analysis re- vealed that EV11 (Table 6, p=0.0005, HR 1.26 (95% Cl, 1.11 to 1.44)) and ARPP19 (p=0.007, HR 2.87 (95% Cl, 1.33 to 6.22)) gene expressions at the diagnosis had significant role in predicting the RFS of the patients. Multivariate Cox’s propor tional hazard model for RFS included diagnosis age, FLT3-1TD status, NPM1 muta tion status and diagnosis phase mRNA expression levels of C1P2A, SET, EV11, WT1, ARPP19, T1PRL and PME1. ln the final model, diagnosis age (p=0.023, HR: 1.07), NPM1 mutation positivity (p=0.048, HR: 0.031 (95% Cl, 0.001 to 0.97)), EV11 (p=0.0005, HR: 1.41), SET (p=0.0097, HR: 0.12 (95% Cl, 0.022 to 0.59)) and ARPP19 (p=0.0001, HR: 58.8 (95% Cl, 7.39 to 467.2)) mRNA expression were in dependent prognostic factors for RFS (Table 7). Added to this, Cox’s typel analy sis revealed that ARPP19 expression (p=0.005) gave additional information in AML patients relapse prognosis after risk group and EV11 mRNA expression were depicted as significant factors in explaining the probability of relapse. Receiver operating characteristic (ROC) analysis also revealed predictive accuracy of ARPP19 detection together with EV11 to be more accurate predictor of relapse than either of the markers alone (EV1 AUC alone 0.68, ARPP19 AUC alone 0.67 and ARPP19+EV11 AUC together 0.77; Figure 6).

Together these results identified high ARPP19 expression as a novel qPCR marker for high relapse risk for patients that are treated with standard in duction therapy lmportantly, the predictive role of ARPP19 was additive when the currently used clinicopathological markers, including risk group classification, were taken into account.

ARPP19 is a novel marker for minimal residual disease (MRD) monitoring

To validate the ARPP19 overexpression in an independent patient co hort, we analysed ARPP19 mRNA expression levels in 48 patients from The Finn ish Hematology Registry and Clinical Biobank (FHRB) (patient cohort2) that had received intensive chemotherapy as an induction therapy, and for which complete clinical data was available. Patient characteristics for these 48 patients are pre sented in Table 4. Median overall survival of the 48 patients was 2.5 years (95% Cl 1.5-3.4) and median follow-up time was 2.5 years (range 7 days - 5.2 years). Gene expression normalization to housekeeping genes and to normal bone mar row control sample was done exactly in the same manner as for patient cohortl.

As a result, ARPP19 was overexpressed in 58% (n=28) of the sample panel, thus providing an independent validation for high rate of overexpression of ARPP19 in a subset of adult AML patients ln this cohort, OS was not in line with risk group classification (data not shown), and therefore the material was not suitable for validation of RFS analysis results observed in cohortl. However, we were able to use samples from cohort2 to address another clinically relevant question whether ARPP19 could be used for disease monitoring during first re- mission phase after intense chemotherapy. Among the cohort2 patients, we could identify all together 9 patients representing the diagnostic phase, first remission phase, and relapse phase, so that each patient had in addition to diagnosis phase sample (n=9), at least either remission (n=4), or relapse (n=8) sample. ARPP19 mRNA expression at the diagnosis phase was higher than in the normal BM sam- ple whereas ARPP19 expression was significantly decreased at remission (Figure 5A). lmportantly, ARPP19 mRNA expression was again increased when patients relapsed, indicating that gradual increase of ARPP19 expression during MRD could be a predictive sign for forthcoming clinical relapse. Difference in the ex pression levels between the diagnosis and remission phase (p=0.003 by Student’s t-test), and between the remission and the relapse (p=0.01 by Student’s t-test) were statistically significant. Very importantly, these conclusions were confirmed in the complete matched set (diagnosis, remission and relapse) of samples from three patients (Figure 5B).

These results both validated marked ARPP19 overexpression in AML, and identified ARPP19 as a potential MRD monitoring marker in AML, wherein increased expression during remission may indicate forthcoming clinical relapse. Thus, ARPP19 may be regarded as a marker of disease activity in AML, and be used as a predictive relapse marker in AML patients having MRD.

Claims

CLA1MS

1. An in vitro method of predicting relapse in a subject having been treated for haematological cancer, wherein the method comprises:

assaying the level of ARPP19 expression in samples obtained from the subject at two or more time points;

detecting a change in the assayed expression levels of ARPP19 be tween the two or more time points;

and predicting the relapse from the change, wherein increased risk of relapse is determined from an increase in the level of ARPP19 expression, or non-increased risk of relapse is determined from a non-increased level of ARPP19 expression.

2. The method according to claim 1, wherein said two or more time points refer to time points after the subject has achieved remission.

3. The method according to claim 1 or 2, wherein said haematological cancer is leukaemia, preferably selected from the group consisting of acute mye loid leukaemia (AML), chronic myeloid leukaemia (CML) and acute promyelocytic leukaemia (APL).

4. The method according to any one of claims 1-3, wherein the expres sion level of ARPP19 is determined at mRNA, cDNA, or protein level.

5. Use of ARPP19 for predicting relapse in a subject having been treat ed for haematological cancer.

6. The use according to claim 5, wherein said haematological cancer is leukaemia, preferably selected from the group consisting of acute myeloid leu kaemia (AML), chronic myeloid leukaemia (CML) and acute promyelocytic leu kaemia (APL).

7. An in vitro method of selecting treatment to a subject diagnosed with AML, the method comprising the steps of:

determining the expression level of ARPP19 in a biological sample ob tained from the subject;

determining whether or not the determined expression level of ARPP19 is increased in the sample by comparing said expression level to a con trol; and

selecting a treatment regime comprising stem cell therapy if said ex pression is increased, or

selecting a treatment regime comprising chemotherapy if said expres- sion is non-increased or decreased.

8. The method according to claim 7, wherein said control is the expres sion level of ARPP19 in a non-diseased sample.

9. The method according to claim 7 or 8, wherein said subject has been stratified into an intermediate risk group on the basis of the European Leuke- miaNet recommendation.

10. A kit for use in the method according to any one of claims 1-4 and 7-9, comprising one or more reagents that specifically detect ARPP19 in a sample obtained from a subject whose relapse is to be predicted or to whom AML treat- ment is to be selected.

11. Use of a kit comprising one or more reagents that specifically de tect ARPP19 for predicting relapse of haematological cancer in a subject accord ing to any one of claims 1-4 or for selecting AML treatment to a subject according to any one of claims 7-10.