WO1994006015A1

WO1994006015A1 - Method to monitor drug therapy and assess metastasis

Info

Publication number: WO1994006015A1
Application number: PCT/GB1992/001581
Authority: WO
Inventors: Simon Peter Robins
Original assignee: The Rowett Research Institute
Priority date: 1992-08-28
Filing date: 1992-08-28
Publication date: 1994-03-17
Also published as: AU2481992A

Abstract

The invention provides a method for detecting the presence of metastasis in subjects who are afflicted with malignancies of non-connective tissues and for monitoring the efficacy of drug protocols. The method is directed to measuring the level of native free crosslinks derived from collagen degradation in biological fluids.

Description

METHOD TO MONITOR DRUG THERAPY AND ASSESS METASTASIS

Technical Field The invention relates to methods of diagnosis in medical and veterinary contexts. More specifically, it concerns methods to assess the effects of drug treatment on connective tissue metabolism and to detect the presence of metastasis using native free crosslinks formed by collagen degradation. Assays for these native free crosslinks in biological fluids, such as urine, provide a useful screening tool and index for the foregoing parameters.

Background Art

Detection of the metastatic state originating from malignant tumors, whatever their tissue of origin, including breast, prostate, thyroid, lung, and kidney, is of importance in designing appropriate therapeutic protocols. Furthermore, it is important to assess the efficacy of protocols employed in attempting to arrest malignancies, including those originating from the foregoing tissues and bone. The present invention is directed to methods to provide such clinical assessments. Published PCT application PCT/GB90/02030, published 11 July 1991 as WO 91/10141, which describes and claims the work of the inventor herein, is incorporated herein by reference. This application describes the use of the level of native free pyridinoline and/or deoxypyridinoline crosslinks as an

SUBSTITUTE SHEET index to metabolic disorders of connective tissue, including bone disease. This PCT application describes in detail the use of either chromatographic techniques or immunoassay to detect and measure levels of such native free crosslinks and establishes a correlation between such crosslinks and diseases which include osteoporosis, Paget's disease, hyperparathyroidism, rheumatoid arthritis, and osteoarthritis. All of these diseases of the bone were found to be characterized by elevated levels of native free crosslinks. Thus, such levels could be used as the basis for screening assays to diagnose these conditions.

It has now been found that levels of native free crosslinks in biological fluids such as urine correlate with the presence of metastases in patients whose condition of metastasis arises not only from bone tumors, but also from malignancies of nonconnective tissue origin. Levels of these native free crosslinks may also be used to monitor the efficacy of therapeutic protocols designed to treat both malignancy and connective tissue disorders.

Disclosure of the Invention

The present invention provides a method to screen patients for the presence of conditions associated with benign and transitional tumors of the bone, such as nonossifying fibroma, enchondroma, osteochondroma, and osteoclastoma, as well as malignant bone tumors such as multiple myeloma, osteosarcoma, chondrosarcoma, Ewing's sarcoma, malignant fibrous histiocytoma; fibrosarcoma, and other metastatic bone tumors. These tumors are metastatic in nature and occur secondary to carcinoma of prostate, breast, thyroid, lung or kidney. Accordingly, detection of these conditions is indicative of a metastatic condition of the tumor of origin.

SUBSTITUTE SHEET Thus, in one aspect, the invention is directed to a method to screen a subject for the presence of a metastatic condition associated with an original tumor, which method comprises measuring the level of native free 3-hydroxypyridinium crosslinks derived from collagen (which crosslinks to not contain peptide chains) in a biological fluid of said subject, and comparing the measured level with the level of said crosslinks in normal subjects. Thus it is possible to assess that a subject having a significantly enhanced level of said crosslinks is likely to have a metastatic condition.

In another aspect, the invention is directed to a method to monitor therapeutic regimens using the level of native free 3-hydroxypyridinium crosslinks in biological fluids measured in a manner similar to that set forth above as an index of efficacy. By taking periodic samples of the biological fluids and comparing the levels of native free crosslinks over time as treatment progresses, the progress of the therapy can be assessed.

Brief Description of the Drawings

Figure 1 shows a chromatographic trace of pyridinoline obtained from an acid hydrolyzate super- imposed on a trace of the pyridinoline obtained without hydrolysis from urine. The figure further compares the elution pattern as determined by fluorescence with the elution pattern as determined by reaction with anti- pyridinoline antibody prepared from hydrolyzate. Figure 2 shows a reverse-phase HPLC of hydrolyzed urine from a healthy subject showing the elution position of total free Pyd and total free Dpd.

Figures 3A and 3B show the Pyd/creatinine and Dpd/creatinine ratios of subject diagnosed for Paget's

SUBSTITUTE SHEET disease, primary hyperthyroidism, and nutritional osteomalacia in comparison to normal individuals.

Figure 4A shows diagrammatically the ratio of total free pyridinium crosslinks to creatinine in subjects with bone metastases and those having no bone etastases. Figure 4B is a similar diagram showing the corresponding Dpd/creatinine ratio.

Figure 5A shows the correlation of the Pyd/creatinine ratio with alkaline phosphatase concentration; Figure 5B shows a similar diagram correlating Dpd/creatinine with alkaline phosphatase.

Figure 6A shows, diagrammatically, the ratio of native free pyridinium crosslinks to creatinine in subjects with bone metastases and those having no bone metastases. Figure 6B is a similar diagram showing the corresponding Dpd/creatinine ratio.

Modes of Carrying Out the Invention

The invention method provides a simple method to assess whether or not a subject known to have a tumor of nonconnective tissue origin is affected by metastases. The ability to make this assessment is critically important in structuring the future therapeutic and lifestyle regimen suitable for the subject. The invention method is straightforward and can be conducted in a noninvasive manner. Specifically, the invention method is directed to measuring the levels of native free collagen-derived 3-hydroxypyridinoline crosslinks to assess metastasis. The levels of these compounds can also be used to assess the progress of therapies designed to treat the metastases or the underlying tumor condition.

Although it is also contained in the above- referenced PCT application incorporated herein, the

SUBSTITUTE SHEET following background information is reproduced here for the convenience of the reader.

Nature of the Crosslinks The abbreviations Dpd and Pyd will be used herein to denote the two known forms of the isolated crosslink itself. Pyd or pyridinoline refers to crosslinks formed wherein the ring N is the € amino of a hydroxylysyl residue; Dpd or deoxypyridinoline refers to crosslinks formed wherein the ring N is e amino of a lysyl residue. (Various methods of denoting these variations have been used; for example, HP has been used to designate the "hydroxylysyl" form, and LP has been used to refer to the "lysyl" form.) Specifically, Dpd represents a compound of the formula:

and Pyd represents a compound of the formula:

SUBSTITUTE SHEET

It is seen that both forms of crosslinks are 1,4,5 trisubstituted 3-hydroxypyridinium residues. Pyd has a free hydroxyl group on the sidechain which can be glycosylated, and it is known to be glycosylated in some tissues. The glycosylation is labile to acid, and also to base, but to a lesser degree. Pyd has been shown to occur as Gal-Pyd; the inventor herein has also demonstrated the presence of Glc.Gal-Pyd in urine (see PCT application WO 89/00715) . These forms of free Pyd have the acetals

respectively, conjugated to the sidechain hydroxyl. It is seen that Dpd contains three chiral centers--those of the three α-amino positions in the sidechains. Pyd contains four such centers, as there is

SUBSTITUTE SHEET an additional chiral center at the sidechain hydroxyl position. Presumably, in unhydrolyzed samples, whether the crosslinks are further derivatized to peptides or not, the three α-amino groups are derived from the natively occurring L-enantiomers, and the OH is in a configuration also determined by the biological system.

A substantial proportion of the crosslinks present in urine (about 40% in adults) is in the form of native "free" crosslinks--i.e. , there are no peptide chains conjugated to the Pyd, glycosylated Pyd, or Dpd structures shown above, even before hydrolysis of the sample is conducted. Thus, by "free" crosslink is meant compounds of the formulas shown above. By "native free" crosslinks is meant the compounds of the formulas shown above as they occur prior to hydrolysis from any peptides which may be conjugated to them.

It is noted with respect to Pyd and Dpd, the chirality of the chiral centers is not specified. It has been shown by the inventor herein that the chirality of the native free crosslinks differs from that which resides in the hydrolyzed forms. Hydrolyzed crosslinks also do not contain sugar residues.

As the native free crosslinks are the product of the biological system, it is assumed that the biologically favored chirality occurs at all three or four chiral centers. Presumably the three chiral centers represented by the α-amino groups of the sidechains are in the L configuration, as in the naturally occurring amino acid, and the chirality of the carbon containing the sidechain hydroxyl in Pyd is also representative of a single configuration. This is confirmed by the results shown in Figure 1, in which the dotted line represents the result of ion-exchange chromatography on sulfonated polystyrene beads (7 μ) equilibrated with sodium citrate performed with the previously isolated Pyd in its native

SUBSTITUTE SHEET free form. As seen in Figure 1, the Pyd isolated directly from urine elutes at a single peak. This is consistent with the presence of only a single diastereomer. After hydrolysis, however, the hydrolyzed free

Pyd elutes as a mixture, shown by the solid line in Figure 1. This is consistent with racemization at the chiral centers to obtain a mixture of diastereomers which no longer exhibit identical chromatographic behavior. Similar results are obtained comparing native free Dpd with hydrolyzed free Dpd.

The "native free" crosslinks thus differ from hydrolyzed free forms of crosslinks. It appears that during conventional acid hydrolysis racemization occurs which changes the configuration of some of the molecules. However, enhancement of the yield of total "native free" crosslinks in the biological sample could also be obtained by proteolytic treatment of total native Dpd and Pyd to liberate the "native free" crosslink form. In addition, the crosslinks per se are identical across species, and other species besides human could be utilized to prepare native free crosslink standards for use in the assay system or for use as immunogens. In particular, porcine urine contains high amounts of native free crosslinks. Any source of the biologically important diastereomer could be used.

Preparation of Antibodies to Native Free Crosslinks Antibodies are prepared to the native free crosslink either as a total fraction or, preferably, to each component of this fraction. Gross separation of the pyridinium linkage in its "free" forms from the fragments containing protein can be achieved, for example, by the method of Fujimoto, D., J Biochem (1983) 9_4:H33-1136 (supra) . In this preparation, a concentrate of urine is

SUBSTITUTE SHEET applied to a Sephadex G-10 column and the total pyridinium-containing fractions eluted. The eluate^* is then applied to a column of phosphocellulose equilibrated with sodium citrate, and eluted with salt. This rather simple procedure results in the "free" crosslinks as a single peak. As the sample is not subjected to hydrolysis conditions this peak contains not only the Dpd and Pyd forms, but also glycosylated Pyd including Gal-Pyd and Glc.Gal-Pyd as described above. Further separation of this native free crosslink fraction is then conveniently conducted by standard methods, for example using ion exchange on sulfonated polystyrene beads as described above, or using HPLC. Typical protocols for this separation are found, for example, in Black, D., et al., Anal Biochem (1988) l££:197-203; Seibel, M.J., et al., J Rheumatol (1989) 16:964-970.

Antibody preparation is by conventional techniques including injection of the mixture or the individual components conjugated to carrier into suitable mammalian subjects such as rabbits or mice according to immunological protocols generally known in the art. The materials are conjugated to carriers such as BSA or tetanus toxoid using standard conjugation methods to enhance immunogenicity. Sera are titrated to determine antibody formation with respect to the immunogen. If desired, spleen cell or peripheral blood lymphocytes may be harvested and immortalized to produce cultures of cells capable of continuous production of monoclonal antibodies immunoreactive with the desired component. These preparations have enhanced specificity with respect to the individual components.

Thus, polyclonal antisera can be obtained which are specifically immunoreactive with the native free form of the crosslinks occurring in biological fluids, in particular in urine. By "specifically immunoreactive" is

SUBSTITUTE SHEET meant that the serum is capable of forming complexes with the native free crosslink forms in the biological fluid with sufficiently greater affinity in comparison to other materials in the fluid to permit determination of the native free forms in an irπmunoassay.

It has been found that antisera prepared from individual animals differ in the range of specificities shown. Polyclonal antisera from some animals immunized with native free crosslink or portions thereof specifically recognize the native free forms in biological fluids to the exclusion of those conjugated with peptides. In addition, it has been found possible to use hydrolyzed free crosslinks as immunogen to obtain such polyclonal antisera. Antisera can readily be screened for those polyclonal antisera which are sufficiently specific for the native free forms, to the exclusion of the peptide conjugated forms to be used in direct immunoassays on the biological fluids. In the alternative, monoclonal antibodies can be prepared from any of the sera and screened for the appropriate specificity.

The availability of routine techniques to obtain monoclonal antibody preparations permits reproducible reproduction of antibodies of the desired specificity. Thus, by utilizing a screening procedure which utilizes as a criterion the ability of the immortalized cell supernatant to immunoreact with, for example, native free Pyd, but to fail to react either with native free Dpd or forms of the crosslinks which are further conjugated to peptides, a reliable source of antibodies which react only with native free Pyd can be obtained. Conversely, it may be advantageous to use, in assessment of biological samples, cocktails of antibodies with these unique specificities so that all native free forms are determined.

SUBSTITUTESHEET Immortalized cell lines which secrete antibodies of the desired specificity can be cultured in vitro for the production of practical quantities of the desired monoclonals using mammalian cell techniques known in the art. Such culture techniques are now available on a commercial scale. In addition, the immortalized cell lines may be injected into mice and a somewhat cruder preparation of the monoclonals isolated as the ascites fluid. The antibody preparation may also be affinity purified if desired using the immunogen as an affinity ligand.

It should be noted that it is unimportant whether the antibodies specific for the native free crosslinks do or do not crossreact with the hydrolyzed form of the crosslink, since the hydrolyzed form of the crosslink is not present in biological fluids.

Conduct of Immunoassays

Accordingly, by utilization of an immunoassay with the antibodies prepared as above it is possible to assay a biological fluid sample without prior fraction- ation or hydrolysis. The specificity for the desired form of native free Pyd or Dpd or both is supplied by the antibody preparation. The immunoassays themselves are conducted using the variety of standard assay protocols generally known in the art. As is generally understood, the assay is constructed so as to rely on the interaction between the specific antibody and the desired analyte for specificity and to utilize some means to detect the complex formed by the analyte and the antibody. The complex formation may be between the antibody itself or an immunologically reactive fragment thereof such as an Fab, Fab' , or F(ab') fragments. The antibody or immunologically reactive fragment thereof may be complexed to solid

SUBSTITUTE SHEET support and used as a capture antibody for the analyte. This protocol may be run in a direct form, wherein the formation of analyte/antibody complex is detected by a fluorescent, radioactive or enzymatic label, or may be run in a competitive format wherein a labeled standard competes with analyte for the antibody. The format may also be constructed as an agglutination assay or the complex may be precipitated by addition of a suitable precipitant to the reaction mixture. The specific design of the immunoassay protocol is open to a wide variety of choice, and the number of clinical assay devices and protocols available in the art is multitudinous.

The antibodies and reagents for the conduct of an immunoassay using standard detection protocols--i.e. , for example radioisotope labeling, fluorescent labeling or ELISA, either in a direct or competitive format can conveniently be supplied as kits which include the necessary components and instructions for the assay.

Since antibodies can be raised specifically to the forms of the native free crosslinks which comprise the various forms thereof, the ratios of these components can be determined as well as their individual levels and their total.

Thus, the assay can be designed to include antibodies or immunologically reactive fragments thereof which will result in determination of total native free crosslinks, or determination of native free Pyd, Dpd, Gal-Pyd, or Glc.Gal-Pyd, or any desired combination thereof. Since the levels of the Pyd and Dpd crosslinks in various tissues can be determined, alteration in their relative amounts can be used as an index for degradation of the particular tissue in question. For example, for most normal adults, the ratio of Pyd/Dpd stays constant throughout adulthood. As bone has a Pyd/Dpd ratio of 4/1 and appears to be the major source of liberated Dpd, an

SUBSTITUTE SHEET elevation in the ratio of Dpd/Pyd may be indicative of bone degradation. (Although aorta also contains Dpd, its turnover rate is low.) Assessment of the level of Dpd in biological fluids also yields a result which is relatively bone-specific. However, it appears that in many instances where a bone disorder is suspected, the total free crosslink level (Dpd + Pyd) can also be used as a measure when additional information is present. When the symptoms do not suggest a disease of cartilage such as rheumatoid arthritis, the majority of the excess crosslink in free form in biological fluids will be, in fact, due to the resorption of bone.

Since other connective tissues, such as cartilage, for the most part contain only Pyd, not Dpd, an elevation in the ratio of Pyd/Dpd may indicate diseases associated with such damage.

In a particularly preferred protocol for conducting the immunoassays of the invention, competition is effected between the 3-hydroxypyridinium in the sample and the corresponding relevant 3-hydroxypyridinium forms coupled to a microtiter plate or other solid support. This method is particularly preferred when a single species of 3-hydroxypyridinium, such as pyridinoline, is to be assessed. The wells of microtiter plates, or suitable portions of other solid substrates, are first coated with the competing 3-hydroxypyridinium moiety, preferably through a specific binding partner coupling reaction. For example, as illustrated hereinbelow, biotin covalently coupled to an easily adsorbed material such as ovalbumin is first adsorbed and coupled to the surface of the plates, and the 3-hydroxypyridinium moiety covalently bound to streptavidin coupled to the adsorbed biotinylated ovalbumin. The microtiter plate or other surface coupled to the competing 3-hydroxypyridinium moiety is then treated in individual portions thereof

SUBSTITUTE SHEET with test samples, standards, and the appropriate antiserum or monoclonal antibody preparation. Antibody is thus coupled to the well or appropriate portion of the solid substrate in an amount inversely related to the amount of competing 3-hydroxypyridinium in the sample. After washing, detection of the coupled antibody is accomplished through conventional techniques.

While immunoassays using the antibodies of the invention are convenient, the native free Pyd and Dpd crosslinks can also be determined in a variety of ways. Since the pyridinoline linkage is fluorescent, direct chro atography of the sample of biological fluid as described in the art can result in separation of Dpd from Pyd and of the glycosylated forms of Pyd and the intensity of the fluorescence of the peaks obtained provides an index to quantitation.

In the methods of the invention, therefore, the native free crosslinks can be determined either as a group or individually by determining the intensity of the fluorescence of the chromatographed material.

As set forth in PCT application WO89/04491 referenced above, the quantity of crosslinks can also be determined using specific electrodes of appropriate redox potential for the ring system. The assays of the present invention, including chromatographic and immunoassays as well as any other methods designed to detect native free crosslinks are useful in screening subjects for the presence or absence of metastatic conditions resulting from malignancies. Malignancies which are known to result in metastatic conditions include breast cancer, prostate cancer, lung cancer and kidney cancer. These result in metastatic conditions that can be classified as benign and transitional tumors, such as nonossifying fibroma, enchondroma, osteochondroma, or osteoclastoma; and

SUBSTITUTE SHEET malignant tumors, such as multiple myeloma, osteosarcoma, chondrosarcoma, Ewing's sarcoma, malignant fibrous histiocytoma; fibrosarcoma; and metastatic bone tumors. Since the condition of the subject can be monitored continuously, application of these assays can also be used to monitor the progress of therapy administered to treat these or other conditions.

In this application, as the therapy is administered, periodic samples are taken of biological fluids, and these are assessed for native free crosslinks. If the therapy is designed to decrease bone resorption, for example, the levels of either of these indicators should decline if the therapy is successful. Conversely, if there is no discernable change or if the levels of native free crosslinks rise, it is clear that the therapy is either ineffective or contraindicated. As set forth above, kits may be constructed which are designed for the conduct of the assays described above in the foregoing context. Suitable kits for detection of "native free" crosslinks will include specific antibodies immunoreactive with these forms, or the immunologically reactive fragments thereof, along with other reagents useful in detecting this interaction. The following examples are intended to illustrate but not to limit the invention.

Example 1 Assay for Native Free Crosslinks in Urine

A. Isolation of U-Pyd and U-Dpd: Urine samples were collected from patients with Paget's disease or hyperparathyroidism (which contains elevated levels of free crosslinks) and from growing children (in which about 10-fold higher concentrations of crosslinks are present compared with normal adults) . After

SUBSTITUTE SHEET concentration 10-fold by rotary evaporation, batches of the urine (20 liters) were subjected to partition chromatography batchwise on cellulose CF1 using butanol:acetic acid:water (4:1:1 v/v/v) as mobile phase. The pyridinium crosslink-containing fraction, eluted from the stationary phase with water, was chromatographβd on a column (3.2 x 150 cm) of Sephadex G-10 eluted with 0.2M acetic acid. Pooled fractions containing the crosslinks were then made 67 mM in Na⁺ and applied to a column (1.7 x 35 cm) of Dowex 50X-X8 ion-exchange resin equilibrated with 67 mM sodium citrate buffer, pH 2.75. After raising the column temperature to 60 C, elution with 67 mM sodium citrate was performed with a linear pH gradient from 2.75 to 5.50 over 500 ml. The column effluent was monitored by fluorescence (ex 325 nm/emm 400 nm) and the pooled fractions containing U-Pyd (364-377 ml) and U-Dpd (397- 416 ml) were desalted by gel filtration on Sephadex G-10 and evaporated to dryness. The yield from 20 liters of urine was 2.5 μmoles U-Pyd and 0.6 μmoles U-Dpd.

B. Results: The isolation procedure set forth in paragraph A of this sample was applied to urine samples from individual patients and the amounts of U-Pyd and U-Dpd were quantitated using fluorescence measurements relative to creatinine as is known in the art (supra) . The values obtained for normal individuals and in 7 patients with bone disorders and arthritic diseases are shown in Table 2. Values are given as the mean ± SEM (n=6 in each group) .

SUBSTITUTE SHEET Table 2

Patient Group U-Pyd U-Dpd (nmol/mmol creatinine)

Normal controls 10.3 ± 1.0 3.27 ± 0.57 Osteoporosis 19.6 ± 2.3 5.90 ± 0.68 Paget's disease 62.5 ± 11.2 19.3 ± 3.83 Hyperparathyroidism 55.9 ± 14.2 16.3 ± 4.81 Rheumatoid arthritis 38.8 ± 8.36 8.92 ± 2.08 Osteoarthritis 25.8 ± 3.22 6.10 ± 0.83

These results show dramatically elevated levels of the free crosslinks in patients known to be suffering from diseases characterized by excessive breakdown of connective tissue.

Table 3 shows the proportions of U-Pyd and U- Dpd as a percentage of the total crosslink measured after hydrolysis in the different patient groups.

s- T T Έ SHi IET Table 3

Patient Group % U-Pvd* % U-Dpd*

Normal controls 43.8 ± 2.5 50.1 ± 5.4 Osteoporosis 41.7 ± 2.0 42.7 ± 2.6 Paget's disease 46.5 ± 2.4 47.4 ± 4.1 Hyperparathyroidism 48.7 ± 6.8 46.2 ± 6.9 Rheumatoid arthritis 38.1 ± 2.6 43.3 ± 1.8 Osteoarthritis 43.4 ± 3.9 47.0 ± 2.2

* Calculated as: (U-Pyd/total Pyd) x 100 and (U-Dpd/total Pyd) x 100. For all groups combined (n=36) , the correlation coefficient between U-Pyd and total Pyd was 0.929 (p<0.0001) and between U-Dpd and total Dpd was 0.952 (p<0.0001).

Since, as shown in Table 3, the percentage of U-Pyd and U-Dpd is relatively unchanged in patients with abnormal conditions as compared to controls, concentra¬ tions of the free crosslinks in urine reflect the same increase in collagen degradation in diseases compared with the controls as do the total crosslinks measured after hydrolysis of the urine. U-Pyd and U-Dpd therefore provide viable indices of collagen degradation to facilitate diagnosis and monitoring of diseases involving abnormalities of connective tissue metabolism.

C. Immunoassay: Native free crosslinks isolated by the method described in paragraph A of this example are used for the preparation of antigen. U-Pyd and U-Dpd are further purified by ion-exchange chromatography with 67 mM-sodium citrate buffer, pH 4.25 using a high-resolution

SUBSTITUTE SHEET resin column of an amino acid analyzer (Locarte Co. Ltd., London, UK) .

For immunization, the isolated crosslinks are covalently attached to bovine serum albumin using carbodiimide reagents and methods well known in the art. Both monoclonal and polyclonal antibodies are raised against the urinary crosslink components. For the production of monoclonal antibodies, Balb/c mice are immunized with urinary crosslink-BSA conjugates, and hybridoma cell lines are prepared using standard techniques after fusions of cells from the spleen or lymph nodes with Ag8 myeloma cells. Polyclonal antibodies are raised in rabbits. Screening of both antisera and hybridoma cell media was performed by ELISA using microtiter plates coated with the appropriate urinary crosslink-gelatin conjugate prepared as described by Robins, Biochem J (1982) 217:617-620.

Assays for each of the crosslink components present in free form in urine are performed by an inhibition ELISA as follows:

Urine samples (5 or 20 μl) or solutions containing 0.2-20 pmol of purified urinary crosslink reference standard are diluted to 110 μl with phosphate buffered saline containing 0.05% Tween-20 detergent (PBS- T) , and are added to 110 μl of primary antibody, immunoreactive fragment, or antiserum diluted 1:5,000 - 1:20,000 in PBS-T. Each sample is prepared in triplicate in round-bottomed, 96-well microtiter plates which are then incubated overnight at room temperature. Portions (200 μl) of the samples are transferred to flat-bottomed microtiter plates previously coated with gelatine conjugate containing the appropriate urinary crosslink component. After 30 minutes, the plate is washed with PBS-T (3 times) and the bound antibodies detected by standard techniques with a biotin-labeled

SUBSTITUTE SHEET antibody prepared against the species of the primary antibody combined with a streptavidin-peroxidase and peroxidase substrate detection system. Color development is measured at 492 nm using an automated microtiter plate reader. Samples containing the analyte decrease the binding of primary antibody to the plate and thus have reduced color concentration. The amount of free crosslinks in the sample is quantified with reference to curves from standards included on each plate computed using log-log plots.

The foregoing assay can be reformatted to be conducted directly by coating the sample suspected of containing antigen in the flat-bottom microtiter plate, and adding labeled primary antibody directly to the wells. After washing, the amount of labeled antibody remaining in the testing solution is determined. A decrease in levels indicates the presence of antigen.

Example 2 sources of Native Free Crosslink

In order to determine a source for native free crosslinks usable as standards in the assays of the invention, the urine of a number of species of large animals was analyzed. In bovine urine, the Pyd/Dpd ratio is 12±2 with only about 15% as free crosslink; the values in sheep are similar except for only about 20-25% is free crosslink. In pig urine, the ratio of Pyd/Dpd is about 5±1 and the proportion of free crosslink relative to total is 42±5%. The concentrations of free crosslinks are about 380 nM for Pyd and 70 nM for Dpd.

Children's urine appears to give a better yield of Dpd than urine from adults. Some preferential loss of Dpd from pig urine occurs when CF-1 cellulose is used in the purification procedure, and overall recovery of Pyd is 40-50% for Pyd but only 20% for U-Dpd. Using

SUBSTITUTE SHEET children's urine as a starting material, recovery for both crosslinks is about 55%.

Accordingly, both children's and pig urine are suitable sources for free crosslink standards.

Example 3 Urinary Crosslinks in Metabolic Bone Diseases

Urine samples were obtained from a total of 47 individuals referred to clinics at Ninewells Hospital, Dundee; the Presbyterian Medical Center, New York; or the Rheumatology Department, City Hospital, Aberdeen. The study included patients with Paget's disease, primary hyperparathyroidism, and osteomalacia. There were 26 Paget's disease patients who had unequivocal radiological evidence of the disease and raised alkaline phosphatase levels. Sixteen patients were in the group who had symptomatic primary hyperparathyroidism as diagnosed by increased levels of parathyroid hormone. Five patients had been diagnosed as having osteomalacia from their history, physical examination and biochemical evaluation.

In the assay, urine samples were frozen within 24 hours of collection and stored at -20°C. Repeat analyses of samples stored for over a year under these conditions showed mean values for both Pyd and Dpd of about 95% or their original value.

Analysis of the crosslinks was performed by HPLC as described by Black et al Anal Biochemistry (1988) 169:197-203 except that heptafluorobutyric acid was used as an ion-pairing agent. Prior to analysis the urine samples (250 μl) were hydrolyzed with an equal volume of concentrated HC1 to release bound forms of the crosslinks.

Urinary hydroxyproline was measured by HPLC or by a commercially available colorimetric method.

SUBSTITUTE SHEET Figure 2 shows a typical HPLC chromatogram of the crosslinks in hydrolyzed urine after prefractionation of the hydrolysate on CF1 cellulose. The two crosslinks, Pyd and Dpd, are well separated from each other and are the major components, as detected by fluorescence monitoring.

The results of the determination of Pyd and Dpd levels (as a ratio to creatinine) for the patients are shown in Figures 3A and 3B, respectively. As shown, for controls the mean values in nm/mm creatinine were in the normal ranges with an average of 25.2 (range 2.1-49.3) for Pyd and 7.2 (range 0.9-14.4) for Dpd. For Paget's disease, the values were Pyd 164.9 ± 119.2, Dpd 80.4 +. 57.4; for primary hyperparathyroidism, Pyd 119.5 ± 61.2, Dpd 52.5 ± 25.9; and for osteomalacia, Pyd 125.1 ± 49.0, Dpd 45.8 ± 18.9. Thus, the average for disease patients is significantly higher than controls.

Example 4 Determination of Crosslinks in the Urine of Metastatic Patients A total of 20 breast cancer patients were studied, 9 with known bone metastases and 11 with no recognized metastases after one year's subsequent follow- up. When the hydrolyzed "total free" crosslinks were determined as described in Example 3, the results showed that seven of the nine patients with metastases had crosslink excretion values higher than those of the control; so did some patients without known metastatic disease. However, there was a clear correlation both of "total free" crosslinks levels with alkaline phosphatase levels. These results are shown in Figures 4A and 4B and 5A and 5B.

Sϋ iS ^■T: S iT i U U ^' T . ϊr. S SH fcS Example 5 Determination of Native Free Crosslinks in the Urine of Metastatic and Non-Metastatic Patients The patients of Example 4 were also studied determining the level of native free crosslinks (using urine without hydrolysis) using HPLC. The native free crosslinks in the urine of these patients also showed a correlation with the metastatic state of the subject. Native free Pyd showed levels of 49.4±12.3 nm/mm creatinine; for the group with no observable bone metastasis, the value was 20.3±2.5 nm Pyd/ m creatinine. Using native free Dpd crosslinks as a measure, the value for the bone metastasis patients was 12.1+2.8 nm/mm, creatinine while for those patients showing no metastasis the value obtained was 5.36±0.83 nm/mm creatinine. The distribution of these values with regard to individual subjects is shown in Figures 6A and 6B.

Example 6 Preparation of Anti-Pyridinoline Antiserum

Pyridinoline was coupled to BSA by the procedure of Robins, S., Biochem J (1982) 207:617-620. Briefly, to 3.1 ml of a solution of 9 mg of bovine serum albumin and 3.8 mg of pyridinoline in 0.1 M MES, pH 5.0 was added an 0.88 ml aqueous solution containing 88 mg 1- (3-dimethylaminopropyl) -3-ethylcarbodiimide (EDCI, Pierce) . After four hours at room temperature, the reaction mixture was exhaustively dialyzed versus phosphate-buffered saline pH 7.0 (PBS). UV and fluorescence measurements indicated 5.8 moles of pyridinoline substituted per mole of albumin.

New England white rabbits were initially immunized by subcutaneous injections at multiple sites with 0.2 mg of the conjugate in 1.0 ml PBS mixed with 1.0 ml of Ribi adjuvant (Ribi ImmunoChemical Research Inc.).

SUBSTITUTE SHEET Subsequent booster immunizations were given at three-week intervals at intramuscular sites. Antisera were collected 10 days after each immunization.

Example 7

Preparation of Pyridinoline-Coated Microplates

A. Biotinylated Ovalbumin:

Ovalbumin was biotinylated by adding 10 mg of biotin-X-2,4-dinitrophenyl-X-L-lysine, succinimidyl ester (Molecular Probes) in 400 microliters of dimethylformamide to a 10-ml solution of PBS containing 150 mg of ovalbumin. After two hours at room temperature, the biotinylated ovalbumin was recovered by G25 column chromatography. Spectrophotometric analysis indicated two biotin per mole of ovalbumin.

B. Pyridinoline/Streptavidin Conjugate:

To a 0.75 ml solution of 5 mg oi: streptavidin in PBS was added 21 μl of dimethylformamide containing 260 μg of N-succinimidyl-3- (2-pyridylthio)propionate (SPDP, Pierce) . After one hour at room temperature the reaction mixture was dialyzed versus PBS. Dithiothreitol was added to a final concentration of 10 mM to reduce the SPDP-streptavidin. After one hour at room temperature, the resulting thiolated streptavidin was purified on a G25 column. To a mixture of 0.5 mg thiolated streptavidin nd 50 μg of pyridinoline in 1.0 ml or PBS was added 180 μg of succinimidyl 4- (N-maleimidomethyl) cyclohexane-1-carboxylate (SMCC, Pierce) in 4 μl of dimethylformamide. After 3 hours at room temperature, the mixture was dialyzed versus PBS. Spectrophotometric analysis indicated between 1 to 2 pyridinolines incorporated per mole of streptavidin.

C. Microplate Coating:

150 microliters of biotin-ovalbumin solutions at 3.8 μg/ml in PBS was added to each well of the

SUBSTITUTE SHEET microplate followed by an overnight incubation at 2-8°C. The microplates were washed with PBS and blocked by adding 200 μl of ovalbumin at 1 mg/ml with an overnight incubation at room temperature. The microplates were then twice washed with PBS, and the streptavidin- pyridinoline conjugate immobilized via the streptavidin- mediated binding to biotin: 150 μl of a solution containing streptavidin-pyridinoline at 100 μg/ml in PBS was added to each well of the biotin-ovalbumin-coated microplate. After one hour at room temperature, the plates were twice washed with PBS. Residual liquid was removed from the microplate by drying overnight in a convection oven at 37°C.

Example 8

Conduct of the Assay Ten μl of urine sample or standards was added to each well of the microplate of Example 7 followed by 150 μl of the antipyridinoline serum at a final dilution of 1:4500 in PBS with 1 mg/ml of BSA and 0.05% Tween-20. the plates were incubated overnight at 2-8°C, then washed three times with 300 μl of PBS. 150 μl of a goat anti- rabbit IgG alkaline phosphatase conjugate (American Qualex) at a final solution of 1:2000 in PBS, 1 mg/ml BSA was dispensed to each well. After one hour at room temperature, the plates were washed three times with 300 μl of PBS. 150 μl of the the enzyme substrate, 2 μg/ml of p-nitrophenylphosphate (Sigma) in 1.0M diethanolamine pH 9.8 was added. After one hour at room temperature 50 μl of 3.ON NaOH was added to each well to stop the enzymatic reaction. The optical density at 405 nm was then measured with a V ax reader (Molecular Devices Corp.) .

SUBSTITUTE SHEET

Claims

1. A method to screen a subject for the presence of metastasis associated with a malignancy, which method comprises measuring the level of native free 3-hydroxypyridinium crosslinks derived from collagen, which crosslinks do not contain peptide chains, in a biological fluid of said subject, and comparing the measured level with the level of said crosslinks in normal subjects so as to assess that a subject having a significantly enhanced level of said crosslinks is likely to have metastasis of said malignancy.

2. A method to monitor a therapeutic protocol directed to a disorder associated with connective tissue or with metastases of malignancies which method comprises repetitively and periodically measuring, during the course of said therapeutic protocol, the level of native free 3-hydroxypyridinium crosslinks derived from collagen, which crosslinks do not contain peptide chains, in a biological fluid of said subject, and comparing the measured level at each measurement with the level of said crosslinks in measurements obtained at previous times during said protocol in said subject so as to assess that a subject having a significantly diminished level of said crosslinks as compared to said earlier measurements is likely to have had a positive response to said protocol.

3. The method of claim 1 or 2 wherein the native free crosslinks are determined as deoxypyridinoline and/or pyridinoline crosslinks.

SUBSTITUTE SHEET

4. The method of claims 1-3 wherein said biological fluid is selected from the group consisting of serum and urine.

5. The method of claim 1 - 4 wherein said malignancy is selected from the group consisting of a malignancy of breast, prostate, thyroid, lung and kidney.

6. The method of claim 1 - 5 wherein said measuring of said native free crosslinks utilizes an immunoassay comprising the steps of: contacting the biological fluid or a 3- hydroxypyridinium crosslink containing fraction thereof with a composition containing an antibody or immunologically reactive fragment thereof which composition is specifically immunoreactive with native free crosslinks under conditions wherein said antibody or fragment immunoreacts with said crosslinks but only minimally immunoreacts with 3-hydroxypyridinium crosslinks containing peptide chains, and detecting the amount of said immunoreaction.

7. A kit for immunoassay determination of the amount or concentration of one or more native free crosslinks in a biological fluid, wherein said kit is used for the determination of the presence or absence of metastasis or the efficacy of a therapeutic protocol, which comprises a set of containers, at least one of which contains a composition containing an antibody or immunologically reactive fragment thereof, said composition being specifically immunoreactive with said native free crosslink, but which antibody only minimally reacts with 3-hydroxypyridinium crosslinks that contain peptide chains, and

SUBSTITUTE SHEET at least one of which containers contains an additional reagent for the conduct of said immunoassay, along with instructions for the conduct of said assay.

SUBSTITUTE SHEET