WO2017060840A1 - Use of htra1 for early detection of colorectal cancer - Google Patents

Abstract

The invention relates to a novel test for the early diagnosis of colorectal cancer (CRC),. The test measures the expression decrease of HtrA1 protein in plasma samples. Its application implies either the use of HtrA1 alone or in combination with the immunochemical fecal occult blood test (iFOBT). The latter application is called "combined test". The new test offers the advantage of affording greater sensitivity than the most reliable available fecal occult blood test (FOBT). It may be hypothesized that the test could be used for routine population-based CRC screening, either alone or combined, before patients who are positive for fecal occult blood are referred for colonoscopy.

Description

USE OF HTRA1 FOR EARLY DETECTION OF COLORECTAL CANCER

Technical field of the invention

The present invention relates to a novel test for the early diagnosis of colorectal cancer (CRC), called "AI.Ma. test". The test measures the expression decrease of HtrA1 protein in plasma samples or histology specimens. It can be used alone or in combination with the immunochemical fecal occult blood test (iFOBT). The latter application is called "combined AI.Ma. test". The new test offers the advantage of affording greater sensitivity than the most reliable available fecal occult blood test (FOBT), the one performed with the immunochemical technique (iFOBT), if used alone on a plasma sample, and a greater sensitivity as well as specificity if combined with the iFOBT. As regards the determination of tumor stage, use of HtrA1 alone has a greater sensitivity than the iFOBT, either according to the TNM classification (TNM staging) and to the Dukes classification. It may be hypothesized that the test could be used for routine population-based CRC screening, either alone or combined, before patients who are positive for fecal occult blood are referred for colonoscopy. State of Prior Art

Colonoscopy is the gold standard examination for the diagnosis of CRC. However its invasive nature, the cost of equipment, and staff training requirements have prevented its diffusion. The first-line screening test for CRC, recommended by the European Council, is based on the fecal occult blood test (FOBT), which in Italy is administered free to individuals aged 50 to 70 years (1 ). The most widely used FOBT is based on a biochemical assay, the guaiac test (gFOBT). Whereas the gFOBT involves dietary restrictions before the procedure, to reduce false positives, a specific diet is not required for an immunochemical test (iFOBT) based on human hemoglobin antibodies, which detect only hemoglobin from the lower gastrointestinal tract. The main weakness of this test is that storage conditions (e.g. excessively high temperature) may increase the rate of false negatives (2). According to the most accurate studies, the iFOBT has a sensitivity of 65.8 % (95 % confidence interval [CI], 55.4-76.3 %) and a specificity of 94.6 % (95 % CI, 94.3- 94.9 %) compared with the gold standard of colonoscopy, whose mean sensitivity and specificity are respectively 88.8 % and 75.4 % (3,4). As regards clinical TNM staging, it has a lower sensitivity for the earlier stages (Dukes' A, 50.0 %; Dukes B, 70 %; Dukes C-D, 78.3 %) (5). Notably, subjects who have continuous bleeding may actually test negative even in the presence of disease. Blood markers may provide a useful alternative to current CRC screening tests, and several studies have suggested markers that could have a greater sensitivity than the iFOBT, like CD26 (sensitivity 81 .8 %, specificity 72.3 %) (6) and tumor pyruvate kinase M2 (sensitivity 79 %, specificity 81 %) (7). Some important studies have integrated and compared new potential markers to the iFOBT (8,9). In a case-control study, Lumachi and co-workers (10) tested five tumor markers - carcinoembryonic antigen (CEA), cancer antigens (CA) 19-9 and 72- 4, cytokeratin fragment (CYFRA) 21 -1 , and osteopontin - both alone and combined. They found that osteopontin provided the highest sensitivity (45.1 %) and CEA the greatest specificity (90.9 %), whereas the combination of all five markers achieved a sensitivity of 74.1 % and a specificity of 94.3 %.

HtrA1 , one of the four members of the HtrA protein family, was first described by Zumbrunn and Trueb (1 1 ). It is a secreted protein found in several normal tissues, including the intestine (12). Several studies have analyzed HtrA1 expression levels in a variety of tumor and control tissue biopsies and have found a correlation among expression level, degree of tumor differentiation, and clinical TNM stage (13).

However, to date HtrA1 protein has never been measured in plasma from subjects with any type of cancer, and in particular no study has been conducted on colorectal cancer. Such earlier studies demonstrate (13) that the levels of HtrA1 tissue expression found in healthy subjects were usually higher than those seen in individuals with a wide range of tumors, even though the difference was not always statistically significant. Moreover, the correlation among HtrA1 tissue expression and histological tumor differentiation, TNM stage, and tumor mass size was contradictory in the different studies. None of these studies have assessed the applicability of plasma HtrA1 to early diagnosis. Moreover, according to the literature, HtrA1 plasma levels have been measured only in healthy pregnant women, thus providing evidence that HtrA1 is also present in plasma (14). However, no studies have measured HtrA1 levels in physiological fluids or liquids from patients with cancer, such as colorectal cancer, as in the present invention.

The purpose of the present invention is thus to provide a new test suitable for colorectal cancer (CRC) screening with improved characteristics compared with the prior art. Summary of the invention

The present invention is based on the discovery that HtrA1 protein present in plasma (as protein concentration, measured as pg/ml), or in histological tissue specimens, (measured as immunostaining intensity %), can usefully be employed for early diagnosis of colorectal cancer.

Therefore, the object of the present application is a diagnostic method of colorectal cancer (CRC) characterized by the fact that it comprises the following steps:

(a) determining the concentration of HtrA1 protein, or a derivative thereof, or another HtrA family member in a biological sample, said biological sample being a biological fluid, and

(b) comparing the concentration measured in the sample to a reference value, wherein measured values lower than the reference value indicate the presence of CRC, wherein the biological specimen is a biological fluid.

In one embodiment of the invention, the method involves, after step (b), an additional step (c), wherein subjects with measured values lower than the reference value are referred for second-level CRC diagnosis and/or CRC treatment.

Advantageously, the second-level diagnosis is diagnostic or therapeutic colonoscopy.

Other objects of the invention are the following embodiments:

A method where the reference value is a threshold concentration (cut-point) (pg/ml) of the protein predetermined experimentally.

A method where the HtrA1 protein derivative is selected from a proteolysis fragment thereof, a variant thereof sharing at least 50 % amino acid identity with HtrA1 protein itself, or a protein comprising the HtrA1 amino acid sequence, or a fragment or variant thereof.

A method where the biological fluid is selected for instance from blood, plasma, serum, urine, saliva, cerebrospinal fluid, and tissue homogenate.

In a preferred embodiment of the invention, the diagnosis is early CRC diagnosis.

In a further embodiment of the invention, the determination of HtrA1 protein, or a derivative thereof, is performed by a technique selected from ELISA, RIA, an immunofluorescence assay, and a Western blotting assay, where preferably the HtrA1 concentration in the biological sample is determined either by comparing the colorimetric, fluorometric, or radioactivity value to a corresponding standard curve, or as the ratio of HtrA1 protein concentration to total protein (pg/ml). An additional object of the invention is the combination of the above described method with another method of CRC diagnosis, preferably a fecal occult blood test (FOBT).

A further object of the invention is a therapeutic approach to CRC that comprises the following steps: a) measuring the concentration of HtrA1 protein, or a derivative thereof, in a biological sample, said sample being a biological fluid;

b) comparing the measured value to a reference value, wherein measured values lower than the reference value indicate the presence of CRC;

c) referring subjects, whose measured values are lower than the reference value, for second-level CRC diagnosis and/or CRC treatment.

Advantageously, step (c) comprises diagnostic or therapeutic colonoscopy. Advantages

The method of the present invention offers significant advantages compared to the various first-level approaches based on the search for fecal occult blood, substantially the iFOBT. It is possible for a lesion to be present but to go undetected by the test, because not all polyps and tumors manifest with bleeding in the early stage, or else bleeding may be intermittent (occurring on some days and not on others).

The limitation of the iFOBT is therefore its low sensitivity, which results in false positives and false negatives. False positive results often involve referral for colonoscopy, an expensive second-line examination and a highly demanding procedure both for the patient and the operators. Subjects with false negative results who escape further examination may have a worse prognosis, because the diagnosis is made at a more advanced stage.

The method of the invention therefore advances the art by providing an alternative and improved technique that is at least as effective as the iFOBT alone. In fact, the AI.Ma. test has a greater sensitivity than the iFOBT, both considering the determination of HtrA1 alone and the combination of the two tests, namely iFOBT and HtrA1 (combined AI.Ma. test). In addition, combination of the two tests involves greater sensitivity as well as specificity than the iFOBT alone. Such characteristics allow a reduction in the number of false negatives as well as false positives, permitting a reduction in the number of subjects being referred for second-level examination, like colonoscopy, and consequently a reduction in healthcare expenditure, as well as a wider coverage of the screening population. In fact, collection of a blood sample is clearly more acceptable to an individual than colonoscopy.

Description of figures and tables

Table 1 : The table shows the results of a sensitivity, specificity, accuracy, positive predictive value (PPV), negative predictive value (NPV), likelihood ratio (LR + and LR-), cut-point, relative to plasma values HtrA1 and feces of 'iFOBT, either alone or in combination.

Table 2: The Table shows the results of a sensitivity, specificity, accuracy relative to the staging of Duke and TNM. The results show the highest sensitivity of the method of the invention on plasma samples compared to iFOBT method.

Figure 1 "Box plots HtrA1 and iFOBT": The figure shows the comparison of HtrA1 dosing in plasma and of iFOBT.

Figure 2: The panel 2a shows the "Roc Curve" of HtrA1 versus iFOBT and 2b panel shows the "Roc Curve" of HtrA1 versus iFOBT and their combination. The results show the highest diagnostic accuracy of the method of the invention or combination with iFOBT method compared to only iFOBT method.

Detailed description of the invention

The method of the invention, also called AI.Ma. test, is a first-line diagnostic test, i.e. one that is easy to administer to large population samples, and is based on the determination of HtrA1 protein concentrations in the biological samples analyzed.

In the most general embodiment, the method envisages the determination of the concentration of HtrA1 protein preferably by a double dosage procedure, in a sample of biological fluid, for instance plasma, using known techniques, and comparison of measured values to a cut-point, used as a reference measure, wherein measured HtrA1 values lower than the reference value (cut-point) will indicate the presence of CRC according to a level of probability assumed a priori.

The cut-point is determined using plasma protein values detected in subjects undergoing diagnostic testing. The Youden index and the Pythagorean theorem (the right-angled triangle and the shortest distance from the vertex) have been used to determine the cut-point.

In a specific embodiment of the invention, the method provides that individuals undergoing the test, whose measured value is lower than the reference value, are referred for subsequent second-line diagnostic examination or directly for CRC treatment. HtrA1

"HtrA1 " refers to the serum protease high temperature requirement A1 in all natural states in which it is found free in biological fluids or tissues. For the purpose of the present application, HtrA1 thus refers to the full-length protein; any protein encompassing the HtrA1 amino acid sequence such as its pre- protein and pro-protein precursors; any proteolytic fragment thereof; any natural variant sharing at least 50 %, 60 %, 70 %, 75 %, 80 %, 85 %, 90 %, 95 %, 98 %, 97 %, 98 %, or 99 % amino acid identity with the HtrA1 protein itself, with the sequence of the respective proteolytic fragment, with any natural derivative produced by post-translation processes or, in tissue samples, with the RNA encoding the protein or its fragments. Each of these elements can be considered as a suitable tumor marker and is thus susceptible of being detected by means of the method of the invention, provided that the respective concentrations precisely reflect HtrA1 levels. Biological samples

The biological samples analyzed by the diagnostic method of the invention are preferably biological fluids such as blood, plasma, serum, urine, saliva, cerebrospinal fluid, and tissue homogenates. For the purpose of the present invention, the expression "biological fluid" is considered to be equivalent to "biological liquid", "physiological fluid" or "physiological liquid".

Methods for HtrA1 detection The HtrA1 concentration found in physiological fluids, tissue homogenates, or histological specimens can be determined by a number of known techniques using commercially available tools that are also known to the skilled person. The preferred method is an ELISA (enzyme-linked immunosorbent assay) test that uses appropriate antibodies capable of selectively binding HtrA1 protein, or a portion or derivative thereof, and of highlighting it with suitable markers, which as a result of enzyme activity generate staining or fluorescence in the test solution.

Absolute HtrA1 concentrations (pg/ml) in the samples analyzed are determined by comparing the colorimetric, fluorometric, or radioactivity values measured in the assay to a corresponding calibration curve or standard curve previously prepared using methods that are known to the skilled person. Alternatively, normalized HtrA1 concentrations are determined as the ratio of HtrA1 protein (pg/ml) to total protein concentration.

Alternative HtrA1 dosing methods include immunofluorescence and Western blotting with densitometric quantification.

The immunofluorescence technique employs antibodies directed against specific antigens, in this case HtrA1 , whose presence is being sought, which are dyed with a fluorescent substance (fluorochrome). Fluorochromes are special substances that, if they are excited by a light beam of suitable wavelength, can emit light at a longer wavelength than the one they have absorbed; thanks to this method it is possible to highlight any antigen-antibody immunocomplexes that are formed.

The main fluorochromes are fluorescein isothiocyanate (FITC) and tetramethyrhodamine (TRITC). The fluorescence observed, which is read under a fluorescence microscope at a wavelength of 500 nm, is yellow-green when fluorescein isothiocyanate is used and orange-red when tetramethyrhodamine is used.

Immunofluorescence assays are currently performed according to two different methods: direct, using a slide, and indirect, using serum.

In the indirect method, which is the technique suitable for the analysis of the present fluid, the patient's plasma is appropriately diluted to obtain the highest test specificity without impairing its sensitivity.

The sample is reacted with the appropriate substrate fixed in the well to detect the protein being sought. Then the preparation is examined under the microscope. Emission of fluorescence indicates the presence of immunocomplexes formed by the antibody and the antigen being sought, in the present case HtrA1 .

Immunostaining method

Paraffin sections are processed for HtrA1 detection as described below.

In brief, sections embedded in paraffin according to standard protocols employed in Pathology departments are deparaffinized and rehydrated in xylene and a graded alcohol series (100 °, 95 °, 75 ° and 50 °). Subsequently, to unmask antigen sites, they are treated with Tween 0.3 % in phosphate buffered saline (PBS) for 25 minutes at room temperature. Sections are subsequently incubated for 30 minutes in 3 % hydrogen peroxide in methanol, to inhibit endogenous peroxidase activity. To block non-specific reactions, sections are incubated in normal goat serum for 45 minutes at room temperature (RT). They are then incubated overnight at 4 °C with a rabbit polyclonal HtrA1 antibody diluted 1 :40 in PBS (Ab38610, Abeam, Cambridge, UK), washed in PBS, and again incubated with biotinylated secondary antibody (Vector Laboratories, Burlingame, CA, USA). The Avidin Biotin Complex peroxidase method (Vector Laboratories), applied for 1 hour at room temperature, and 3,3'-diaminobenzidine hydrochloride used as a chromogen (Sigma, St. Louis, MO, USA) enable HtrA1 localization in the tissue being analyzed.

Next, sections are stained with Mayer's hematoxylin, dehydrated, and mounted in Eukitt solution (Kindler GmbH and Co., Freiburg, Germany) for light microscopic observation. Negative controls are performed by omitting the primary or the secondary antibody. A further negative control is performed by using an isotype control antibody (rabbit IgG: Ab27478, Abeam) according to the same method used for the primary antibody.

Another technique that can be used in the present invention is western blotting associated to densitometric quantification.

Western blotting, or immunodetection, is an immunochemical technique enabling identification of a given protein in a protein mix separated by electrophoresis through recognition by suitable antibodies. Proteins are then placed on a support, commonly a nitrocellulose membrane, for the immunomediated reaction and enzymatic detection of the antigen-antibody complex. In an earlier study (14), HtrA1 protein was identified by Western blotting and quantified by densitometry using the Chemidoc tool and Quantity- One software of Bio-Rad Laboratories (Milano, Italy). The Western blotting technique highlights, without any doubt, in a highly specific way, the protein found in the sample, and in fact is often employed as a test of specificity in immunological reactions.

The RIA (radioimmunoassay) technique can also be used in the method of the invention like a common ELISA, where the colorimetric or fluorometric signal is replaced by a radioactive signal. The technique, albeit sensitive, is less preferred because it is more complex to use as well as more expensive, due to the steep cost of the radioactive markers.

Even though the method of the invention can effectively be employed alone in CRC diagnosis, in a specific embodiment it is used in combination with another CRC diagnostic approach.

Preferably said other method is a fecal occult blood test (FOBT).

The FOBT can be performed by different techniques, namely immunochemical assays (iFOBT) or biochemical tests (gFOBT). Alternatively, the determination of fecal occult blood can be performed by measuring fecal porphyrin (HemoQuant test). iFOBT

The immunochemical determination (iFOBT) requires no special preparatory diet. The test is based on human hemoglobin antibodies, which detect only hemoglobin from the lower gastrointestinal tract. The main weakness of this test is that storage conditions (for instance excessively high temperature) may increase false negatives (1 ). Alternative methods to dose fecal occult blood

gFOBT

The gFOBT is the most widely used FOBT and is based on a biochemical test (the guaiac test). However, the gFOBT requires dietary restrictions before the procedure, to reduce false positives.

Stool examination with guaiac involves spreading some stool on absorbent paper that has previously been treated with a chemical substance. Hydrogen peroxide drops are then dispensed on the paper, and if traces of blood are present the paper changes color within 1 -2 seconds. The method is based on the fact that the heme of hemoglobin has an effect similar to peroxidase, involving fast rupture of the hydrogen peroxide molecule. In some situations, like gastric bleeding or upper proximal intestinal tract bleeding, the guaiac method may be more sensitive than the tests based on globin detection, because globin is destroyed in the upper intestinal tract to a greater extent than is heme. Several gFOBT kits with different sensitivity are available on the market. Only the high-sensitivity tests are recommended for colorectal cancer screening. The good clinical performance of the guaiac stool test is dependent on adequate dietary preparation.

Quantification of fecal porphyrin

The HemoQuant test, unlike the gFOBT, affords accurate hemoglobin quantification and has been validated for the analysis of gastric acid and urine, besides feces. The heme portion of intact hemoglobin is converted chemically to protoporphyrin by oxalic acid and ferrous oxalate or ferrous sulfate and the porphyrin content of the sample, before and after hemoglobin conversion to porphyrin, is quantified by comparative fluorescence against a reference standard. The specificity for hemoglobin is enhanced by subtracting the fluorescence measured in a blank sample added with citric acid, to offset the possible confounding effect due to the presence of some non-specific substances. Accurate measurement and quantification have proved highly valuable in several clinical investigations. Fecal DNA Test

The PREGEN-Plus test extracts human DNA from the stool sample and can detect changes that may be associated to cancer. The test checks 23 single DNA alterations, of which 21 are point mutations of genes APC, KRAS and p53, and examines the BAT26 gene, which is involved in microsatellite instability (MSI).

The method of the present invention is effectively employed for CRC diagnosis. The results of clinical studies, reported in the experimental section below, have demonstrated that the method being claimed is highly effective in early CRC diagnosis, namely stage A according to Dukes staging or stages T1 and T2 and N1 and N2 according to TNM staging. The presence of disease is demonstrated by the reduction, in the fluid or tissue analyzed, of HtrA1 levels below an experimentally identified cut-point threshold. The following cut-points were calculated in relation to the Dukes classification: stage A, 6.83 ± 3 pg/ml, stages B and C, 8.85 ± 3 pg/ml. The following cut- points were estimated in relation the TNM stage classification: stages T1 and T2, 6.83 ± 3 pg/ml; stages T3 and T4, 10.07 ± 3 pg/ml; stages N1 and N2, 6.83 ± 3 pg/ml; stages N3 and N4, 8.85 ± 3 pg/ml.

A specific embodiment of the invention is described below with all relevant experimental data.

Confirmatory diagnostic methods

After first-line examination with the AI.Ma. test, involving detection of HtrA1 levels, there may follow a second phase if values lower than the reference threshold value are found in a subject. In such case, the subject with CRC diagnosis is referred for a second-line examination, for confirmation or exclusion of the diagnosis made by the method of the invention. The most reliable confirmatory examination is substantially colonoscopy or an equivalent procedure. Colonoscopy has a dual function, diagnostic and therapeutic, because the visual inspection is supported by tools that enable mechanical removal of cancerous or pre-cancerous lesions. Experimental data from 40 patients have demonstrated that the colonoscopy findings matched precisely those obtained using the method of the invention.

EXPERIMENTAL SECTION

Plasma HtrA1 was dosed in CRC patients and control subjects using a commercial ELISA kit (Uscn Life Science Inc., Wuhan, PRC) which envisages the dosage of biological fluids including serum, plasma, and other fluids.

Total protein concentration was determined using the Bradford assay (Bio-Rad Laboratories). The final results were expressed as the ratio of HtrA1 to total protein concentration (pg/ml).

The subjects undergoing plasma testing also provided a stool sample where fecal occult blood was dosed with the iFOBT. All patients gave their written informed consent to be included in the study (as per art. 170 bis of the Italian Penal Code).

Recruitment of patients and controls.

Patients with a diagnosis of primary CRC confirmed by histology were enrolled consecutively. The choice of this recruitment modality was based on practical as well as economic reasons. Controls. The inclusion criteria were as follows: (1 ) hospitalization due to conditions unrelated to CRC, such as trauma, tonsillectomy / adenoidectomy, varicose veins, upper respiratory infection; (2) residence in Ancona province; (3) comparable age (+ 5 years). Two control subjects were paired to each case, matching the patient in age and province of residence. All participants were Caucasian subjects of Italian nationality. Exclusion criteria for all participants were: a family history of CRC, intestinal polyps, cancer, diabetes mellitus, obesity, gastric ulcer, alcoholism, and anal fistulae or fissures. Additional exclusion criteria for cases were earlier adjuvant treatment (chemotherapy or radiotherapy) and cancer surgery. Plasma HtrA1 was determined in CRC patients and control subjects using a commercial ELISA kit (Uscn Life Science) that envisages dosage of biological fluids such as serum, plasma, and other fluids.

Statistical and epidemiological analyses. As in a previous study of plasma conducted by our group (14), the normality of HtrA1 protein distribution in plasma was previously tested with Shapiro-Wilk test. Continuous variables were expressed as mean + standard deviation (sd) and ordinal variables as median and interquartile range. The relative risk of CRC associated to HtrA1 levels was calculated by deriving odds ratios (OR) from the maximum likelihood estimates and the 95 % CI based on the exponentiation of the logistic regression beta parameter. The differences in plasma HtrA1 and iFOBT values between cases and controls were assessed with the Wilcoxon or the chi-square test, as appropriate. TNM differences were evaluated with the Kruskal-Wallis test. Test accuracy was assessed by estimating the receiver operating characteristic (ROC) curves both for the AI.Ma. test and for the iFOBT. The parameters that are commonly used to evaluate diagnostic test performance, i.e. sensitivity, specificity, accuracy, positive predictive value (PPV), negative predictive value (NPV), positive likelihood ratio (LR+), and negative likelihood ratio (LR-), were calculated for the HtrA1 and the iFOBT as well as for the combination of HtrA1 + iFOBT. Sensitivity, specificity, and accuracy and the respective ROC curves were calculated for the Dukes and TMN staging. The accuracy of estimates was evaluated with 95 % CI. A value of p < 0.05 was considered statistically significant. All statistical analyses were carried out using SAS, STATA, and R software.

The results of the statistical and epidemiological analyses of the parameters analyzed for the HtrA1 and the iFOBT are summarized in Tables 1 and 2, respectively. Their graphical representations are reported in Figures 1 and 2, respectively.

Clinical Study

Type and number of patients recruited for the AI.Ma test:

Each sample was paired to 2 healthy controls.

N=40: patients with CRC (cases).

N=80: controls (healthy controls). Methods used to recruit patients and controls

Patients with a diagnosis of primary CRC confirmed by histology were consecutively recruited at Azienda Ospedaliero Universitaria Ospedali Riuniti, Ancona (Italy).

Control individuals were consecutively recruited from the patients admitted to the Internal Medicine and Surgery departments of the same institution. Inclusion criteria for the recruitment of control subjects were as follows: (1 ) hospitalization due to conditions unrelated to CRC, like trauma, tonsillectomy / adenoidectomy, varicose veins, and upper respiratory infection; (2) residence in Ancona province; (3) comparable age (± 5 years). Each case was paired to two control subjects who matched the case in terms of age (± 5 years) and area of residence. All participants were Caucasian subjects of Italian nationality. For all participants, exclusion criteria were: a family history of CRC or intestinal polyps, malignancy, diabetes mellitus, obesity, gastric ulcer, alcoholism, and anal fistulae or fissures. Further exclusion criteria for cases were previous adjuvant treatment (chemotherapy or radiotherapy) and cancer surgery.

Type of samples analyzed: Plasma, feces.

Methods used for laboratory analyses

Plasma HtrA1 was analyzed using an ELISA assay (Uscn, Life Science Inc., Wuhan, PRC).

Feces were analyzed using the OC-SENSOR DIANA iFOB Test (Eiken

Chemical Co., Tokyo, Japan). The latter is a fully automated quantitative immunostaining method based on latex agglutination turbidimetry, whose clinical cut-off is 100 ng/ml hemoglobin.

ELISA test Whole venous blood (ca. 3 ml) was collected in EDTA tubes and centrifuged at 1500 x g for 15 minutes. To measure plasma HtrA1 concentrations, samples were diluted 1 : 100, and two 100 μΙ aliquots per case were analyzed by ELISA according to the manufacturer's instructions. Negative and positive internal controls are provided with the kit. Total protein concentration was determined with the Bradford assay (Bio-Rad Laboratories, Milano, Italy). The results were calculated as the ratio of HtrA1 (pg/ml) to total protein concentration (pg/ml). The plasma aliquots were stored at - 80 °C until use.

This ELISA kit uses the Sandwich enzyme immunoassay. The ELISA kit plate was coated with a specific HtrA1 antibody. Plasma aliquots (100 μΙ) were dispensed into the wells of the ELISA microplate, leading to binding of the HtrA1 protein found in the sample to the antibody. After 2 hour incubation at 37 °C, plasma was removed. Then, 100 μΙ of a biotinylated detection antibody specific for HtrA1 was added to each well. After 1 hour incubation, again at 37 °C, wells were washed 3 times with 350 μΙ of rinsing solution (provided with the kit). Subsequently, 100 μΙ of an avidin-horseradish peroxidase (HRP) solution was added to each well, incubated for 30 minutes at 37 °C, and washed 3 times with 350 μΙ of rinsing solution provided with the kit. Then, 90 μΙ of solution containing the substrate was added to each well. The samples containing HtrA1 turned blue. After 20 minute incubation at 37 °C in the dark, the enzyme- substrate reaction was completed by adding 50 μΙ of a sulfuric acid solution, which made the well content turn yellow. Optical density (OD) was measured spectrophotometrically at a wavelength of 450 nm ± 2 nm. The absorbance value is proportional to the HtrA1 concentration. The HtrA1 concentration in samples was calculated by comparing the OD of samples to the values of the standard curve that had been prepared simultaneously with the samples, in the same plate, employing an HtrA1 concentration range from 0 to 2,000 pg/ml and using the standard protein according to the manufacturer's instructions. Immunochemical Fecal Occult Blood Test (iFOBT)

Each participant was given a stool collection kit (Eiken Chemical Co., Tokyo, Japan) complete with instructions. Stool samples were collected using the applicator attached to the bottle cap; the test tube was then closed and shaken. The bottle is designed to contain ca. 10 mg of feces and 2.0 ml of buffer.

The kits were collected and stored in a refrigerator at 4 °C. Samples were analyzed by the iFOBT and the results expressed as hemoglobin concentration.

Statistical analysis Statistical and epidemiological analysis. As in our previous studies the normality of HtrA1 distribution was tested in advance using the Shapiro-Wilk test (14).

Continuous variables are presented as mean + standard deviation (sd) and ordinal variables as median and interquartile range.

The relative risk of CRC associated with HtrA1 levels was calculated by deriving the OR (odds ratio) maximum likelihood estimates and their 95% CI by exponentiation of the logistic regression beta-parameter.

The differences in plasma HtrA1 and iFOBT values between cases and controls were evaluated using Wilcoxon's test or the chi-square test as appropriate. Differences in TNM stage were subjected to Kruskal-Wallis test.

To test the accuracy of the diagnostic tests used to discriminate between two conditions, which in our study are referred to as cases (subjects with cancer) and healthy controls, receiver operating characteristic (ROC) curves were plotted for HtrA1 and FOB values.

We also calculated the proportion of positive and negative test results in relation to disease diagnosis (positive predictive value, PPV and negative predictive value: NPV, respectively).

Both the HtrA1 , which for iFOBT, which for the combination HtrA1 ~ iFOBT, the following parameters were evaluated to interpret the accuracy of a diagnostic test: positive predictive value (PPV) and negative (VPN), LR + and LR -

(likelihood ratio). The same parameters were also calculated for the staging of

Duke and TNM, with the respective ROC curves

The precision of the estimates was expressed as 95 % CI.

P values < 0.05 were considered statistically significant. All statistical analyses were carried out using SAS, STATA and R software.

Results

The results observed in patients with CRC and of healthy individuals are shown in Tables 1 and 2 and illustrated in Figure 2. The concentration levels of HtrA1 in the plasma of 40 patients with colorectal cancer were significantly lower than the 80 controls (mean + sd: 5.2 + 2.6 vs 16.2 + 14.2, p< 0.0001 ); it is interesting to note that in the CCR cases you have a lower concentration of HtrA1 and higher levels of FOB.

The dell'HtrAI overall performance and iFOBT as a diagnostic test are represented by the ROC curves (Figure 2). The area under the curve (AUC- area under curve) was for HtrA1 85.0% (CI 78.3-91 .7%), while for the iFOBT the area under the curve was 77.4% (C! 63.0-86.4 %), being the latter much lower than all'HtrAI . The next step was to assess the overall performance of the combination of the two HtrA1 + iFOBT test. This latter result has shown a considerable increase of the area under the curve with a value of 91 .6% (C.1.85.7-97.4%).

Regarding the use as a diagnostic test detects: with a cut-point of 6.83 (Youden's index = 0.56) to HtrA1 you have a sensitivity SE = 75.0% (CI 67.1 - 82.9%) and a specificity SP = 81 .3% (CI 74.1 -8.4%), a positive predictive value PPV = 66.7% (C! 52.9-80.4%) and a negative predictive value NPV = 86.7% (C! 79,0-94.4). A cut-point of 138.5 (Youden's index = 0.52) to iFOBT clearly demonstrates lower levels of SE = 64.9% (56.1 -73.6% C. I.) and VPN = 75.9% (C. I. 64.5-87.3).

Furthermore, estimates of SE and SP of 2 combined markers give an SE = 73.0% (CI 64.9-81 .1 %) and SP = 97.9% (CI 95.2-100.0%), with a PPV = 96.4% (CI 89.6-100.0) and VPN = 82.1 % (CI 72.1 -92.2) with a Youden index of 0.71 (the best accuracy), indicating that the combined test results (Ai. Ma. test) a higher specificity and a very high sensitivity to the nearest 98%, compared to the gold standard iFOBT that was found to have a specificity of 87.2.

This result is particularly relevant because it indicates that there is a reduction of more than 10% of false positives, subjects which in clinical practice would be to undergo colonoscopy, with also cost-saving. So the AI.Ma. tests could be much more cost-effective.

Finally, as shown in Table 2, the HtrA1 levels measured in plasma are also more sensitive and specific for tumor stage compared with the gold standard of the iFOBT, even when the determination of HtrA1 is not combined with the iFOBT. This is relevant to clinical and therapeutic practice, since a relationship may be hypothesized between HtrA1 level and response to treatment, with clear implications for prognosis.

REFERENCES

Altobelli E, Lattanzi A, Paduano R, Varassi G, di O o F: Colorectal cancer prevention in Europe: Burden of disease and status of screening programs. Prev Med 2014 Feb 14;62C: 132-141 .

Grazzini G, Ventura L, Zappa M. et al. Influence of seasonal variations in ambient temperatures on performance of immunochemical faecal occult blood test for colorectal cancer screening: observational study from the Florence district. Gut 2010 59, 151 1-1515. Sawbridge D, Probert C. Population-based screening in colorectal cancer current practice and future developments: faecal biomarkers review. J Gastrointestin Liver Dis. 2014 Jun;23(2): 195-202. lumb AA, Halligan S, Pendse DA, Taylor SA, Mallett S. Sensitivity and specificity of CT colonography for the detection of colonic neoplasia a_er positive faecal occult blood testing: systematic review and metaanalysis. Eur Radiol 2014; May;24(5): 1049-58 Morikawa T, Kato J, Yamaji Y, Wada R, Mitsushima T, Shiratori Y. A comparison of the immunochemical fecal occult blood test and total colonoscopy in the asymptomatic population. Gastroenterology. 2005 Aug; 129(2):422-8. De Chiara L, Rodriguez-Pineiro AM, Rodriguez-Berrocal FJ, Cordero OJ, Martinez-Ares D, Paez de la Cadena M. Serum CD26 is related to histopathological polyp traits and behaves as a marker for colorectal cancer and advanced adenomas. BMC Cancer. 2010; 10:333 Li R, Liu J, Xue H, Huang G. Diagnostic value of fecal tumor M2-pyruvate kinase for CRC screening: a systematic review and meta-analysis. Int J Cancer. 2012 Oct 15; 131 (8): 1837-45. doi: 10.1002/ijc.27442. Tao S; Haug U; Kuhn K; Brenner H. Comparison and combination of blood-based inflammatory markers with faecal occult blood tests for noninvasive colorectal cancer screening. Br J Cancer 2012; 106(8): 1424-30. Lumachi F, Marino F, Orlando R, Chiara GB, Basso SM. Simultaneous multianalyte immunoassay measurement of five serum tumor markers in the detection of colorectal cancer. Anticancer Res. 2012 Mar;32(3):985- 8. Zumbrunn J and Trueb B: Primary structure of a putative serine protease specific for IGF-binding proteins. FEBS Lett 398: 187-192, 1996. De Luca A, De Falco M, Severino A, Campioni M, Santini D, Baldi F, Paggi MG, Baldi A. Distribution of the serine protease HtrA1 in normal human tissues. J Histochem Cytochem.2003 Oct;51 (10): 1279-84. Altobelli E, Marzioni D, Lattanzi A, Angeletti P M. Htr1 : Its future potential as a novel biomarker for cancer. Oncology Report, Epub: 28.05.2015; 34: 555-566. D. Marzioni, T. Lorenzi, E. Altobelli , S.R. Giannubilo , F. Paolinelli, C. Tersigni, C. Crescimanno, V. Monsurro, A.L. Tranquilli, N. Di Simone, M. Castellucci. Alteration of maternal plasma HTRA1 level in preeclapsia complicated by IUGR. Placenta 2012, 33: 1036-1038.

T. Lorenzi a, D. Marzioni , S. Giannubilo , A. Quaranta , C. Crescimanno, A. De Luca , A. Baldi , T. Todros , A.L. Tranquilli , M. Castellucci. Expression patterns of two serine protease HtrA1 Forms in human placentas complicated by preeclampsia with and without intrauterine growth restriction. Placenta 2009; 30: 35-40.

Claims

1 . A method for the diagnosis of colorectal cancer (CRC) characterized by the fact that it comprises the following steps:

a. determining the concentration of HtrA1 protein, or a derivative thereof, in a biological sample, said biological sample being a biological fluid,

b. comparing the measured HtrA1 value to a reference value, wherein measured values lower than the reference value indicate the presence of CRC.

2. A method according to claim 1 , wherein said HtrA1 reference value is a threshold concentration (cut-point) (pg/ml) predetermined experimentally.

3. A method according to claim 1 , wherein said HtrA1 protein derivative is selected from a proteolytic fragment thereof, a variant thereof sharing at least 50 % amino acid identity with the HtrA1 protein, or a protein comprising the amino acid sequence of HtrA1 , or a fragment or variant thereof.

4. A method according to claims 1 and 2, wherein said biological fluid is selected from blood, plasma, serum, urine, or saliva.

5. A method according to claims 1 to 3, wherein said diagnosis is early CRC diagnosis.

6. A method according to claim 5, wherein early diagnosis is performed at stage A of the Dukes stage-classification or at stages T1 and T2 or N1 and N2 of the TNM stage-classification.

7. A method according to any one of claims 1 to 6, wherein the cut-point concentrations that indicate the presence of CRC according to the Dukes classification and the TNM classification are given by the following values:

Early stage Dukes' A: 6.83 ± 3 pg/ml,

Stages Dukes' B and C: 8.85 ± 3 pg/ml,

Stages TNM T1 and T2: 6.83 ± 3 pg/ml,

Stages TNM T3 and T4: 10.07± 3 pg/ml,

Stages TNM N1 and N2: 6.83 ± 3 pg/ml,

Stages TNM N3 and N4: 8.85 ± 3 pg/ml.

8. A method according to any one of claims 1 to 7, wherein HtrA1 protein or its derivative is determined using a technique selected from ELISA, RIA, immunofluorescence, and Western blotting.

9. A method according to any one of claims 1 to 8, wherein the HtrA1 concentration in the biological sample is determined either by comparing colorimetric, fluorometric, or radioactivity values to a corresponding standard curve, or as the ratio of HtrA1 to total protein concentration (pg/ml).

10. A method according to any one of claims 1 to 9, that is combined with another first-line approach to CRC diagnosis.

1 1 . A method according to claim 10, wherein said further first-line diagnostic approach is a fecal occult blood test (FOBT).

12. A method according to any one of claims 1 to 1 1 , comprising after step (b) an additional step (c), wherein any subjects whose measured HtrA1 values are lower than the reference value are referred for second-line CRC diagnosis and/or CRC treatment.

13. A method according to claim 12, wherein step (c) involves performing diagnostic and/or therapeutic colonoscopy.

14. A method of treating CRC comprising the following steps:

a. Determining the concentration of HtrA1 protein, or a derivative thereof, in a biological sample, said biological sample being a biological fluid,

b. comparing the measured HtrA1 value to a reference value, where measured values lower than the reference value indicate the presence of CRC,

c. referring individuals presenting measured HtrA1 values lower than the reference value for second-line CRC diagnosis and/or CRC treatment.

15. A therapeutic method according to claim 14, wherein step (c) involves diagnostic and/or therapeutic colonoscopy.