DIAGNOSIS OF CANCER

Abstract

The data provided herewith show that, biomarkers found within the cells of a tumor or cancer may also be present at adequate levels within the patient's urine to enable detection. Method of diagnosis, patient selection and treatment are provided, along with kits and devices.

Claims

1. A method of determining whether a patient does, or does not, have cancer, comprising determining the presence or absence of an oncoprotein in a sample of bodily fluid from the patient, wherein the presence of the oncoprotein in the sample indicates that the patient has cancer.

2. The method of claim 1 wherein the oncoprotein is an intracellular protein, such as PRL3, PRL1, VHZ, c-myc, H-ras, AKT-1, p53, Rac1, FAK, Runx1, Estrogen Receptor (ER), PTEN, b-actin or GAPDH, preferably PRL3.

3. The method of claim 1 wherein the sample of bodily fluid from the patient is a urine sample.

4. The method of claim 1 wherein the oncoprotein is an extracellular oncoprotein, such as Her2, N-Cadherin, PDGF receptor (alpha), FLT-3 or p-EGFR.

5. The method of any one of the preceding claims further comprising quantifying the level of oncoprotein in the sample.

6. The method of any one of the preceding claims wherein the cancer is an initial or early stage cancer.

7. The method of any one of the preceding claims wherein the cancer is selected from bladder, lung, breast, stomach, nasopharangeal or prostate cancer.

8. The method of any one of the preceding claims wherein the oncoprotein is soluble oncoprotein.

9. A method of determining whether a patient does, or does not, have cancer, the method comprising determining the presence or absence of an oncoprotein in a sample containing exosomes obtained from a bodily fluid from a patient, the sample having been enriched for exosomes, and/or containing substantially only exosomes, wherein preferably the sample is, or is derived from, urine.

10. The method of any one of the preceding claims wherein the oncoprotein is present within the exosomes in the sample.

11. A method of selecting a patient for an anti-cancer therapy, the method comprising detecting the presence of an oncoprotein in a sample of, or derived from, bodily fluid from the patient, and selecting a treatment based on the oncoprotein determined to be present, where the bodily fluid is preferably urine.

12. A lateral flow device comprising one or more anti-oncoprotein antibodies.

13. A kit for detecting oncoprotein in a sample of bodily fluid, the kit optionally including the lateral flow device of claim 12.

14. A method for diagnosing bladder cancer in a patient, the method comprising isolating exosomes from a urine sample from the patient and determining the presence of oncoprotein in the exosomes, wherein the presence of oncoprotein in the exosomes in indicative that the patient has bladder cancer.

15. An anti-cancer therapy antibody for use in the treatment of cancer in a patient, wherein the patient has been selected for treatment by a method comprising determining, in a sample of bodily fluid from the patient, the presence of an oncoprotein.

16. The method, kit or lateral flow device according to any one of the preceding claims wherein the oncoprotein is PRL3.

17. A method of treatment, the method comprising treating a patient selected by a method according to any one of claims 1 to 10.

18. The method of treatment according to claim 17, comprising administering anti-cancer therapy to the patient.

19. The method according to any one of claim 11, 15 or 18 wherein the anti-cancer therapy is an antibody therapy.

20. The method according to claim 19 wherein the antibody therapy is anti-PRL3 antibody therapy.

Description

FIGURES

[0226] FIG. 1: PRL-3 can be used as a biomarker for cancer stages. High PRL-3 expresses in 8/10 later stage of patients, while in 1/10 early stage of patients.

[0227] FIG. 2: Oncoprotein expression in two normal urine samples and six urine samples from bladder cancer patients. Oncoproteins are commonly detectable in cancer urine samples.

[0228] FIG. 3: Lanes 1-2 normal urines, 3-8 cancer urines. PRL-3 can be used as a biomarker for late cancer stages. High PRL-3 expresses in 8/10 later stage of patients, while in 1/10 early stage of patients. Exosome associated PRL3 is present in the urine of bladder cancer patients. Purified exosome fractions from bladder cancer patient urine samples were analysed with antibodies against PRL3, CD63, exosome marker.

[0229] FIG. 4: 34 urine samples from lung cancer patients.

[0230] FIG. 5: 34 Urine samples from lung cancer patients.

[0231] FIG. 6: Oncoprotein expression in lung, urines samples from breast, stomach and nasopharageal carcinoma patients.

[0232] FIG. 7: 22 urine samples from lung cancer, cystitis, prostatic Adenocaronima v benign prostatic hyperplasia patients.

[0233] FIG. 8: Comparing with normal urine (lane 1), we detect super-strong PRL-3 expressing levels in a large B cell lymphoma patient (lane 2) and low levels in lung cancer urines (lane 3-4).

[0234] FIG. 9: PRL-3 negative samples (1,2,3,5) and positive samples (8,9,12,14,16) on a dot blot (left panel) are reconfirmed by a western blot (right panel). We may be able to use dot blot to test large samples at one time.

[0235] FIG. 10: Antibody purity tested by SDS PAGE. From the left Ab223, 318, 1992, 1994, reference MoAb and reference rabbit antibody (4%-20% Tris-Glycine Gel)

[0236] FIG. 11: Dose response curve for PRL3 prototype device.

[0237] FIG. 12: A QC colour chart used to score test line signal; B Prototype test device; C Test device interpretation; D Prototype test device using negative, positive and diluted positive urine samples.

[0238] FIG. 13: Intracellular PRL-3 oncoprotein can be secreted into cell culture media and is present in 61% of cancer urines, but not in normal urines. (a) Western blotting of PRL-3 in matched lysates and conditioned culture media of the indicated GC cell lines. CANX, calnexin. (b) Summary of % PRL-3 positivity in urine samples from all cancer patients and normal individuals studied. (c-f) Western blot detection for PRL-3 in the urines of (c) normal individuals and GC patients, (d) nasopharyngeal cancer patients and (e) bladder cancer patients. Representative blots are shown. Mr, relative molecular mass (kDa).

[0239] FIG. 14: Effective anti-PRL3 antibody treatment results in a loss of urinary PRL-3, and mechanistically involves intra-tumoral accumulation and recruitment of immune effectors. (a) Western blotting for PRL-3 protein in matched urine and tumor samples from untreated or anti-PRL3 antibody-treated mice harboring PRL-3+ SNU484 or PRL-3− MKN45 orthotopic gastric tumors. Upper panels, excised stomachs at Day 28 (SNU-484) or Day 56 (MKN45). (b) Proposed mechanism of action of anti-PRL3 antibody on PRL-3+ cancer cells. Two forms of secreted PRL-3 antigen are depicted—a soluble, free form, or bound to exosomes.

EXAMPLES

Example 1

[0240] We examined intracellular oncoproteins: PRL-3, PRL-1, VHZ, c-myc, H-ras, AKT-1, p53, Rac1, FAK, Runx1, Estrogen Receptor, PTEN, b-actin, GAPDH from 101 cancer urines (26 bladders, 44 lungs, 10 breasts, 6 Stomach, 15 NPC), and 11 normal urines. We often detect these intracellular oncoproteins in cancer urines. We also detect mutant forms of GAPDH and b-Actin in cancer (but not normal) urines. The results are show in FIGS. 1-8.

[0241] We examined extracellular oncoproteins: Her2, N-Cadherin, PDGF receptor (Alpha), FLT-3, p-EGFR from 81 cancer urines (6 bladders, 44 lungs, 10 breasts, 6 Stomach, 15 NPC), and 11 normal urines, we only detected FLT-3 at low levels in 3 out of 6 bladder cancer urines using western blot.

[0242] We detected intracellular oncoproteins (much more frequently) than extracellular oncoproteins in cancer urines but not in normal urines. Without wishing to be bound by theory this suggests that extracellular oncoproteins could be more tightly associated with lipid bilayers, therefore, they are more difficult to dissociate from the cell membrane and to travel with exosomes into urine. However, intracellular oncoproteins can be more easily and freely formed and secreted via exosomes to travel into body fluids (blood, urines, saliva). Our findings will facilitate diagnosis for cancer biomarkers used for cancer treatments.

Example 2

Exosome Isolation and Kits

[0243] This example provides a sample protocol for obtaining proteins from exosomes in a bodily fluid sample. The protocol is represented in FIG. 9.

[0244] Protocol: Cat No. EXOAB-KIT-1

[0245] If samples are frozen, thaw on ice

[0246] Combine 10 ml samples+2 ml ExoQuick-TC i.e, (if 500 ul of samples then 100 ul of ExoQuick-TC)

[0247] Mix well by inversion three times

[0248] Place at 4° C. for overnight

[0249] Centrifuge at 1500*g for 30 minutes or (1.5 rfc)

[0250] Remove supernatant, keep Exsosome pellet

[0251] Centrifuge at 1500*g for 5 minutes to remove all trace of fluid

[0252] Add 100 ul RIPA (radio immunoprecipitation assay) buffer or PBS (according to pellet) to exosome pellet and vortex briefly for 15 seconds.

[0253] Place in the room temperature for 20 minutes and take the protein reading at 595 nm.

[0254] Add SDS dye and heat up the samples for 5 minutes at 100° C.

[0255] Perform the Western blot.

Example 3

Evaluation of Prototype Rapid Test Device.

[0256] I. Purpose: [0257] 1. Study the feasibility to develop a prototype rapid test device for the detection of biomarker PRL3 in patient urine samples.

[0258] II. Background

[0259] Bio-marker PRL3 was found by Dr. Zeng Qi of IMCB of National University of Singapore in 1998. It had been showed its elevated level in thousands of cancer patient. Dr. Zeng had developed two monoclonal and two rabbit antibodies anti-PRL3. This study intends to evaluate these four antibodies and to investigate the feasibility to use these antibodies to develop an immunochromatographic device using lateral flow technology to distinguish the urine between cancer and normal person.

[0260] Rapid test format (lateral flow device) has the advantage of short assay time (only 10-20 minutes to see the test results). Assay also can be performed at the sense or doctor office without special equipment. Dr. Zeng has the interesting to develop such rapid test and contracted AD Consultant (Phoenix, Ariz.) to perform the feasibility study by using these antibodies. The followings are the results of this study.

[0261] III. Recommendations:

[0262] Based on conclusions in part IV, we recommend: [0263] Set up the specification [0264] 1. Sensitivity, to be determined [0265] 2. Specificity, to be determined [0266] 3. Limit of Detection (LOD): to be determined. [0267] 4. 15 minute assay [0268] 5. Sample type: urine sample [0269] 6. Storage: room temperature. [0270] 7. Expiration date: 24 months after manufacturing [0271] 8. Interference study: to be determined [0272] 9. Submission type: to be determined [0273] 10. Time schedule: to be determined [0274] Use MoAb 223 as captured antibody (for coating) and rabbit anti-PRL3 (1992) as detector (conjugated to gold particles) for rapid test. [0275] Use Nitrocellulose membrane AE99 as solid phase to bind the antibody 223. [0276] Run the clinical samples (50 positives and 100 negatives) to determined specificity and sensitivity of this prototype [0277] If needed, the further improvements will include increasing sensitivity and specificity achieved by using affinity purified rabbit anti-PRL3 antibody. [0278] Feedback for further modification if needed.

[0279] IV. Conclusions: [0280] 1. MoAb 223 and polyclonal antibody 1992 is the best pair to develop PRL3 rapid test. MoAb 223 was used as detector and polyclonal 1992 as capturer. (see V. 2) [0281] 2. The detectability of 10-20 ng PRL3-GST/ml can be achieved in 15 minutes (see V. 3.3) [0282] 3. A response curve from 12.5 ng/ml to 4,000 ng/ml was shown in V. 3.1 [0283] 4. The negative urine and low level sample reproducibility was shown in V. 3.2 [0284] 5. This prototype can distinguish one positive urine (B) from other 8 negative urine sample (see 3.4) [0285] 6. Antibody purification was in V.1 [0286] 7. Antibody inventory was in VI.

[0287] V. Test Results and Discussions:

[0288] 1. Antibody Purification

[0289] Ascites of hybridoma 223 and 318 and antiserum of 1992 and 1993 were received on May 28, 2014. All the four antibodies were purified using Protein. The results were summarized in the Table.

TABLE-US-00003 Summary of Protein G Purification of Monoclonal and Polyclonal Antibodies Antibody yield Volume, Protein G Purified from Antiserum Antiserum/Ascites ml Antibody or Ascites Lot Number MoAb 318, 5.1 3.42 mg/ml × 3 ml = 10.26 mg/5.1 mg = R-1405001 ascites 10.26 mg 2.0 mg/ml MoAb 223, 2.3 5.96 mg/ml × 0.82 m = 5.6 mg/2.3 ml = R-1405002-1 ascites 4.88 mg 2.4 mg/ml 2.06 mg/ml × 0.35 ml = R-1405002-2 0.72 mg Rabbit anti-PRL3 5.5 ml 4.45 mg/ml × 14.3 ml = 63.6 mg/5.5 ml = R-1406002 Antiserum #1992 63.6 mg 11.56 mg/ml Rabbit anti-PRL3   6 ml 3.54 mg/ml × 17.8 ml = 63 mg/6 ml = R-1406003 Antiserum #1993 63 mg 10.44 mg/ml

[0290] The purity of antibody was tested by SDS-PAGE, see FIG. 10

[0291] Discussion: [0292] 1. Only light chain (25 kDa) and heavy chain (55 kDa) were seen in PAGE. Purity of antibodies is satisfied. [0293] 2. The antibody recovery yield from ascites is 2.0-2.5 mg/ml. This was calculated after dialysis and concentration and based on small amount of ascites (5.1 ml and 2.3 ml). [0294] 3. The antibody recovery yield from two rabbit antiserum was 10 to 11 mg/ml respectively.

[0295] 2. Study on the Best Matching Pair

TABLE-US-00004 The Designed Chart for Antibody Pairing Au-Ab Conjugate MoAb MoAb Rabbit anti- Rabbit anti- 223 318 PRL3 #1992 PRL3 #1993 Coating MoAb 223 1 2 3 4 Antibody MoAb 318 5 6 7 8 Rabbit anti- 9 10 11 12 PRL3 #1992 Rabbit anti- 13 14 15 16 PRL3 #1993

[0296] Sixteen pairs were studied and the results were listed in the following two Tables.

TABLE-US-00005 The Negative Sample Testing Result Antibody Sample Pairing # buffer 6190 Urine 1 Urine 2 dH.sub.2O PBS 1 0 0 0 0 0 2 0 0 0 0 0 3 0 0 0 0 0 4 0 0 0 0 0 5 0 0 0 0 0 6 0 0 0 0 0 7 0 0 0 0 0 8 0 0 0 0 0 9 3 0 0 0 0 10 0 0 0 0 0 11 3 0 0 0 2 12 0 0 0 0 0 13 0 0 0 0 0 14 0 0 0 0 0 15 0 0 3 0 0 16 0 0 3 0 0

TABLE-US-00006 The Positive Control Testing Result PRL3 GST Antibody Pairing 100 ng/ml 1 2 3 0 4 6 4 3 5 2 6 3 7 3 8 2 9 3 10 0 11 8 12 3 13 2 14 3 15 6 16 6

[0297] Result and discussion: [0298] 2.1 From the positive control testing result, rabbit antibody 1992 showed the highest affinity (pairing 11). [0299] 2.2 Based on information of pairing 1-8, 9, 10, 13 and 14, it inferred the affinity of MoAb223 is higher than 318. [0300] 2.3 Without affinity purification, polyclonal antibody usually may contain some antibodies came from the infection during animal immunization. The good example is infected by Pseudomonas. After regular purification, this antibody will be introduced into the test line and conjugation in the test device, therefore, to cause false positive if patient sample contains Pseudomonas. [0301] 2.4 Due to the reason in 2.3, this study will not consider using polyclonal antibody simultaneously as detector and capturer in the test device. [0302] 2.5 Affinity purified rabbit anti-PRL3 can be an option if the sensitivity and specificity are not satisfied after the prototype trial.

[0303] 3. The Performance of Prototype PRL3 Rapid Test

[0304] To perform the PRL3 prototype device on dose response curve and reproducibility, this study used a rapid test reader to quantify the test line signal into digital.

[0305] 3.1 PRL3 Response Curve

[0306] In this study, PRL3 antigen was diluted in sample buffer 6129 (in house). 140 ul of diluted PRL3 was applied to the sample well of the test cassettes (device). After 15 minutes, the test device was read by a rapid test reader. Triplicates were run for each PRL3 level. CV % was calculated for each level.

[0307] Result and Discussion:

TABLE-US-00007 Conc. Run ng/ml 1 2 3 Mean CV % 12.5 0.0087 0.0083 0.0086 0.00857 2.4 25 0.0132 0.0123 0.0127 0.012733 3.5 50 0.0181 0.0186 0.0232 0.019967 14.0 100 0.0425 0.0481 0.0497 0.046767 8.0 200 0.1002 0.1069 0.1054 0.104167 3.3 400 0.1820 0.1925 0.2200 0.198233 9.8 1000 0.4686 0.5232 0.4998 0.4972 5.5 2000 0.8349 0.7645 0.7000 0.766467 8.8 4000 1.2854 N/A N/A 1.2854 N/A

[0308] Discussion:

[0309] 3.1.1 After 2000 ng/ml, the curve became less deep (less sensitivity).

[0310] 3.1.2 Not hook affection was seen below 4000 ng/ml. The analyst in the curve was PRL3-GST. It is unknown how free PRL3 will be in the hook affection.

[0311] 3.2 Reproducibility

[0312] Assay protocol was same as described in 3.1. 20 runs were repeated for a negative urine sample defined and provided by IMCB. Ten runs were repeated for a low positive PRL3 prepared in sample buffer. CV % was calculated each.

TABLE-US-00008 20 Readings of Negative Urine #2 Reading Sample number = 10 Mean CV % 0.0023, 0.0014, 0.0033, 0.0013, 0.0022 0.0027 40 0.0024, 0.0054, 0.0024, 0.0019, 0.0014, 0.0030. 0.0026, 0.0025, 0.0018, 0.0041 0.0035, 0.0036, 0.0021, 0.0049, 0.0020

TABLE-US-00009 10 Readings of One Weak Positive Control Reading Sample number = 10 Mean CV % 0.0210, 0.0196, 0.0164, 0.0171, 0.0202, 0.0170 14.1 0.0163, 0.0160, 0.0148, 0.0128, 0.0163

[0313] Discussion

[0314] 3.2.1 The CV for negative urine is 40%, but the highest and lowest reading was 0.0054 and 0.0014. The threshold for this reader is 0.010 or higher as positive. So this data is acceptable.

[0315] 3.2.2 The low positive CV is 14.1% and acceptable.

[0316] 3.3 The Limit of Detection (LOD)

[0317] A QC Control Color Chart was used to score test line signal. Score 3 is a borderline and under 3 will regard as negative.

[0318] PRL3-GST was diluted in sample buffer at each level shown on the table.

[0319] 100 ul of dilution was applied to the sample well of test cassette. Score the signal at 15.sup.th minute.

TABLE-US-00010 ng/ml 10,000 5,000 1,000 500 100 50 20 10 score 10 9 8 7 6 5 4 3 +/− + + + + + + + +

[0320] Discussion

[0321] 3.3.1 The detectability was 10-20 mg T3-GST/ml.

[0322] 3.2.2 Not hook affection was seen at as high as 10,000 ng/ml.

[0323] 3.4 Clinical Urine Sample Study

[0324] Five negative urine samples and two positive urine samples were received from IMCB. Additional three urine samples were collected from our lab. One of the positive urine samples (L) was used to qualify antibodies. So, only one positive urine sample (Urine B) was used in the final testing using prototype device.

[0325] 3.4.1 Assay Protocol

[0326] 1). Bring the urine samples to the room temperature (caution: PRL3 has a different migration behavior on the membrane when it is cool, so bring the sample to room temperature before run the test is very important).

[0327] 2). Apply 100 ul of urine sample to the test device.

[0328] 3). Read the result at 15th minute.

[0329] 4). Test result interpretation:

[0330] 3.4.2 Assay Results

[0331] Sample #1to #5 were negative urine from IMCB, #6 to #8 were from our lab, #9 was positive urine B. Urine B was positive and 1:10 dilution was also seen weak positive.

TABLE-US-00011 #1 #2 #3 #4 #5 #6 #7 #8 #9 − − − − − − − − +

[0332] VI. Antibody Inventory

TABLE-US-00012 Concentration, Volume Total Antibody/lot number mg/ml ml Mg MoAb 223/ R1405002-1 5.96 mg/ml 0.39 ml 2.3 mg +) R1405002-2 2.06 mg/ml 0.35 ml 0.72 mg 3.0 mg MoAb 318 3.42 mg/ml  2.8 ml 9.5 mg R1405001 Rabbit anti-PRL3 1992 4.45 mg/ml   14 ml 62.3 mg R1406002 Rabbit anti-PRL3 1993 3.54 mg/ml 17.4 ml 61.5 mg R1406003

Example 4

[0333] Unexpectedly, we found secreted PRL-3 oncoprotein in 61% of multiple types of human cancer urines and in all cancer urines derived from PRL-3+ (but not PRL-3−) tumor-bearing mice. Urinary PRL-3 levels were significantly reduced after effective treatment with PRL3-antibody. We propose urinary PRL-3 as a novel diagnostic and a surrogate biomarker for therapeutic response monitoring of PRL3-antibody therapy in multiple cancer types.

[0334] Materials and Methods

[0335] Western blotting. 200 μg of lysates were resolved in separate wells of 12% SDS-polyacrylamide gels and transferred to nitrocellulose membranes before blocking and probing with the indicated primary antibodies at a 1:1,000 dilution overnight at 4° C. After thorough washing with TBS-T buffer (20 mM Tris pH 7.6, 140 mM NaCl, 0.2% Tween-20), the membrane was incubated with the respective horseradish peroxidase (HRP)-conjugated secondary antibodies at a 1:5,000 dilution for 1 h, washed with TBS-T, and visualized using a chemiluminescent substrate (Pierce).

[0336] Antibodies. B cell marker (CD45/CD220, clone RA3-6B2), NK cell marker (CD335/Nkp46, clone 29A1.4), and HSP70 (cat #610607) antibodies were purchased from BD Pharmingen. Calnexin (cat #2679) antibody was purchased from Cell Signaling. CD63 (cat #sc-15363) antibody was purchased from Santa Cruz Biotechnoloy. GAPDH (clone MAB374) antibody was purchased from Millipore.

[0337] Urine analysis. Morning urine samples, collected with patients' consent from the National University Hospital of Singapore, were stored at −80° C. prior to analysis. Samples were centrifuged twice at 4,000×g for 20 min at 4° C. to remove insoluble debris prior to analysis. Equal volumes (30 μL) of urine samples were loaded onto gels for western blotting.

[0338] Exosome isolation. Exponentially-growing cells at 70-80% confluence were washed twice with PBS prior to incubation with 10 ml of serum-free RMPI-1640 for 24 h. Conditioned medium was centrifuged twice at 300×g for 10 min to remove dead cells and debris. Pre-cleared conditioned medium was subsequently concentrated to ˜500 μL using an ultra-centrifugal 3K filter concentrator (Milipore), and exosomes were extracted from the retentate using the Total Exosome Isolation Reagent (Invitrogen) according to the manufacturer's instructions. For human urine exosome extraction, 1 mL of pre-cleared urine was directly used for exosome extraction as described.

[0339] Results Intracellular PRL-3 oncoprotein can be secreted into cell culture media and is present in 61% of cancer urines, but not in normal urines. In 2008, we first reported untraditional immunotherapies against intracellular oncoproteins and demonstrated that PRL-3 antibody could be taken up by tumor cells.sub.22. However, it was still unclear how and where antibody recognition of intracellular antigens took place. Herein, we found an unrecognized natural phenomenon that “intracellular” PRL-3 protein could be detected in concentrated culture media from corresponding PRL-3+, but not PRL-3−, cancer cell lines in vitro (FIG. 13a, lanes 5-8). As a control, the ER-localized protein calnexin (CANX) was exclusively found in lysates but not in conditioned media (FIG. 13a), thus ruling out non-specific contamination by dead cells or cellular debris. Since PRL-3 has promising cancer biomarker potential based on microarray and histological studies.sub.6, we proceeded to investigate if “secreted” PRL-3 might have clinical relevance as a biomarker by analyzing urine samples from both healthy individuals and cancer patients.

[0340] A total of 15 urine samples from healthy individuals and 195 urine samples from cancer patients were analyzed by western blot to detect PRL-3 protein. Unexpectedly, PRL-3 was readily detected in an average of 61% (119 out of 195) of urine samples from patients with different types of cancer (FIG. 13b), but not in any normal urine samples (FIG. 13c, lanes 1-15). Specifically, urinary PRL-3 protein was detected in 10/12 (83%) of gastric cancer urines (FIG. 13c, lanes 16-27), 12/17 (70%) of nasopharyngeal cancer urines (FIG. 13d), 30/67 (45%) of bladder cancer urines (FIG. 13e), 56/85 (66%) of lung cancer urines (FIG. 13f), 8/10 (80%) of breast cancer urines, and 3/4 (75%) of prostate cancer urines (data not shown). Our results from these 210 urine samples identify PRL-3 as a common cancer-specific urinary protein.

[0341] Since PRL-3 protein does not have a sequence peptide for classical secretion via the ER/Golgi pathway, we considered that it might be secreted via non-classical exosome secretion. Exosomes are cell-membrane and/or endosomal-derived vesicles between 50 and 150 nm present in many biological fluids and cell culture media.sub.33. To determine if secreted PRL-3 might be exosomal in nature, we performed exosome fractionation of urine samples from patients with different types of cancer, using tetraspanin CD63 as a control exosomal protein.sub.34. Surprisingly, we detected exosome-associated PRL-3 exclusively in bladder cancer urines (FIG. 3) but not from other types of cancer urines (data not shown). Since exosomes secreted by epithelial bladder cancer cells bypass physical exclusion limits enforced by glomerular filtration.sub.35 and can directly enter the bladder's urine pool, our results indirectly indicate that PRL-3 can be secreted from tumor cells in at least two forms in vivo: 1) As a ˜20 kDa soluble, filterable form presented in multiple types of cancer urines. This ‘free PRL-3’ may leak out into body fluids by tumor necrosis, apoptosis, or tumor cell lysis. Since ‘free PRL-3’ is small enough (˜20 kDa) to pass through the kidney filtration system, which has an estimated cut-off of 70 kDa.sub.35, it can readily accumulate in cancer urines. 2) As an exosome-associated form exclusively found in urines of bladder cancer patients. Since circulating exosomes from other cancer tissues (such as gastric, liver, lung) cannot pass through glomerular filtration, only bladder cancer cells with unhindered access to the bladder urinary system could shed such PRL-3-containing exosomes directly into urine. In summary, secreted PRL-3 is a cancer-specific marker comprising of at least two forms—a soluble, ‘free’ form (detectable in urine from various cancer patients), and a larger, exosome-associated form (detectable in urine of bladder cancers patients only).

[0342] Urinary PRL-3 is a novel surrogate biomarker for therapeutic response monitoring of anti-PRL3 antibody therapy. Since PRL-3 could be frequently detected in urine samples from cancer patients, we questioned if urinary PRL-3 expression was reflective of the presence of genuine PRL-3+ tumors in vivo. Due to the difficulty in obtaining clinical matched tumor-urine samples to validate this relationship, we instead used PRL-3+ SNU-484 and PRL-3— MKN45 orthotopic gastric mouse models to compare the expression of PRL-3 in matched tumor-urine pairs. In addition, each orthotopic model was sub-divided into 2 groups—mice receiving PBS (untreated), or anti-PRL3 antibody treatment (treated)—to elucidate the relationship between anti-PRL3 antibody therapy and urinary PRL-3 expression.

[0343] In untreated PRL-3+ SNU-484 tumor-bearing mice, PRL-3 was highly abundant in urine (FIG. 14a, odd lanes 1-9). However, urinary PRL-3 was no longer detectable in all mice after anti-PRL3 antibody treatment, in line with a decrease in intratumoral expression of PRL-3 (FIG. 14a, even lanes 2-10). Importantly, the loss of urinary PRL-3 signal from anti-PRL3 antibody treated mice corresponded with stomach tumor shrinkage in each case (FIG. 14a, upper panels), suggesting that urinary PRL-3 could be useful as a surrogate biomarker of anti-PRL3 antibody therapeutic efficacy. As a control, we did not detect urinary PRL-3 in mice carrying PRL-3-MKN45 orthotopic tumors, regardless of anti-PRL3 antibody therapy (FIG. 14a, lanes 11-12). Thus, urinary PRL-3 is specifically detected in mice carrying PRL-3+ but not PRL-3− cancers, and diminishes upon treatment with anti-PRL3 antibody.

[0344] Discussion

[0345] This study identifies the biomarker potential of secreted urinary PRL-3 for diagnostic and therapeutic response monitoring. We detected urinary PRL-3 in an average of 61% of multiple human cancer patients.

[0346] Although soluble PRL-3 was detected in urines from multiple cancer patients, we detected exosome-associated PRL-3 only in the urines of bladder cancer patients, but not from patients with other malignancies. The likely explanation of this urine is the physical constraint imposed by glomerular filtration, which only allows passage of proteins smaller than 70 kDa from the plasma into the Bowman's capsule for urinary excretion.sub.35. Thus soluble, ‘free’ PRL-3 (˜20 kDa), but not exosome-associated PRL-3, remained detectable in the urine from these other cancer types. Nonetheless, the presence of exosome-associated PRL-3 presents an intriguing possibility where budding exosomes from PRL-3+ tumors could serve as anchor points within tumor areas for anti-PRL3 antibody recognition in vivo and initiation of an effector immune response (FIG. 14b).

[0347] The close correlation between tumor and urinary PRL-3 expression observed in mouse models suggests that urinary PRL-3 expression could be used as a prospective diagnostic biomarker for PRL-3-targeted cancer therapies (including anti-PRL3 antibody) in a variety of human malignancies. In addition, urinary PRL-3 could also function as a surrogate biomarker, providing a non-invasive, fast, and simple qualitative method for clinicians to infer therapeutic efficacy.