WO2002037113A2

WO2002037113A2 - Clinical and functional validation of targets from disseminated cancer cells

Info

Publication number: WO2002037113A2
Application number: PCT/EP2001/012786
Authority: WO
Inventors: Michael Giesing; Hansjörg GRILL; Beatrix Boeckmann; Bernhard Suchy
Original assignee: Individual
Current assignee: Individual
Priority date: 2000-11-03
Filing date: 2001-11-05
Publication date: 2002-05-10
Anticipated expiration: 2003-05-03
Also published as: AU2002216009A1; WO2002037113A3

Abstract

The present invention relates to method for the cancer-related clinical validation of a target from disseminated cancer cells, to methods for the finctional validation of the targets, to merhods for treating cancer taking accounts of the status of at least one of said validated targets and to kits for target-related cancer treatment.

Description

Clinical and functional validation of targets from disseminated cancer cells

The present invention relates to methods for the cancer-related clinical validation of a target from disseminated cancer cells, to methods for the functional validation of said targets, to methods for treating cancer taking account of the status of at least one of said validated targets and to kits for target-related cancer treatment. Building on this, target-related development of active substances is possible.

The spreading of solid tumors has been known for a long time. Cancer cells enter the bloodstream in particular during the course of vascularization of a solid tumor. They circulate in the bloodstream and likewise in other body fluids and may form metastases according to their molecular endowment.

Reports on the relevance of disseminated cancer cells for the clinical management of malignant diseases are still controversial.

For example, several studies regarding cytokeratin (CK) 20 mRNA expression in blood indicate that the detection of disseminated cancer cells by a CK20-specific RT PCR assay is related to the stage of disease (Fujii Y et al., Jpn J Cancer Res 1999 Jul; 90 (7): 753-757; Denis MG et al., Int J Cancer 1997 Oct 21 ; 74 (5): 540-544; Soeth E et al., Cancer Res 1997 Aug 1 ; 57 (15): 3106-3110; Soeth E et al., Int J Cancer 1996 Aug 22; 69 (4): 278-282; Gudemann C J et al., J Urology 2000 Aug; 164(2): 532-536) and that detection of epithelial cells in preoperative blood was associated with reduced disease-free and overall survival (Hardingham JE et al., Int J Cancer 2000 Jan 20; 89 (1) :8-13). However, others report that the clinical significance of detecting occult cancer in peripheral blood remains to be determined (Majima T et al., Jpn J Clin Oncol 2000 Nov; 30 (11) : 499-503), CK20-mRNA was an useful marker to detect circulating cancer cells in breast cancer but there was no correlation between CK-20 mRNA expression and stage or auxiliary lymph node status (Bae JW et al, J Korean Med Sci 2000 Apr; 15 (2) : 194-198) or there was no correlation between the detection of CK20 mRNA in the peripheral blood and disease progression and survival in patients with known metastatic colorectal cancer (Wyld DK et al., Int J Cancer 1998 Jun 19; 79 (3): 288-292).

The situation for CK19 mRNA is similar (Denis MG et al., Int J Cancer 1997 Oct 21 ; 74 (5) : 540-544; Nakamura T et al., Anticancer Res 2000 Nov; 20 (6C) : 4739-4744; Majima T et al., Jpn J Clin Oncol 2000 Nov; 30 (11) : 499-503; Smith BM et al., J Clin Oncol 2000 Apr; 18(7): 1432-1439; Berois N et al., Eur J Cancer 2000 Apr; 36(6): 717-723; Kahn HJ et al., Breast Cancer Res Treat 2000 Mar; 60(2): 143-151 ; Yeh KH et al., Anticancer Res 1998 Mar; 18 (2B): 1283-1286; Bae JW et al, J Korean Med Sci 2000 Apr; 15 (2) : 194-198; Lin JC et al., Kaohsiung J Med Sci 2000 Jan; 16 (1 ) : 1- 8).

Another example for controversial discussion on the relevance of disseminated cancer cells is the detection of tyrosinase mRNA in melanoma patients. While several studies suggest a lower disease free survival for melanoma patients in which disseminated cancer cells have been detected by means of tyrosinase mRNA (Curry BJ et al., J Clin Oncol 1999 Aug; 17 (8) : 2562-2571; Battayani Z et al., Arch Dermatol 1995 Apr; 131 (4) : 443-447; Schrader AJ et al., Anticancer Res 2000 Sep; 20 (5B) : 3619-3624; Mellado B et al., Clin Cancer Res 1999 Jul; 5 (7) : 1843-1848 Mellado B et al., J Clin Oncol 1996 Jul, 14 (7) : 2091-2097 ; Kunter U et al., J Natl Cancer Inst 1996 May 1 ; 88 (9) : 590-594; Baldi A et al., Anticancer Res 2000 Sep; 20(5C): 3923-3928; Schrader AJ et al., Melanoma Res 2000 Aug; 10(4): 355-362), others mention low detection rates which correlate poorly with the clinical stage of melanoma suggesting that tyrosinase mRNA may be of limited value in the management of malignant melanoma (Alao JP et al., Melanoma Res 1999 Aug; 9 (4) :395-399; Jung FA et al., J Clin Oncol 1997 Aug; 15 (8) : 2826-2831).

Another prominent example for the current dilemma is prostate cancer and the expression of prostate-specific antigen (PSA). While there are reports confirming the potential utility of this assay in identifying patients more likely to have recurrences (de la Taille et al., Int J Cancer 1999 Aug 20; 84 (4) : 360-364; Gao CL et al., J Urol 1999 Apr; 161 (4) : 1070-1076; Wood DP jr et al., J Clin Oncol 1997 Dec; 15 (12) : 3451- 3457) others deny a relationship between PSA mRNA positivity, pathological and clinical features (Llanes L et al., BJU Int 2000 Dec; 86 (9) :1023-1027; Oefelein MG et al., J Urol 1999 Aug; 162 (2) : 307-310).

Furthermore, the expression of α-fetoprotein (AFP) fails to correlate with disease stage (Kienle P et al., Arch Surg 2000 Feb; 135 (2) : 213-218; Hautkappe AL et al., Cancer Res 2000 Jun 15; 60(12): 3170-3174; Barbu V et al., Hepatology 1997 Nov; 26(5): 1171-1175) but seems to correlate with recurrence in patients with hepatocellular carcinoma (Wong IH et al., Clin Cancer Res 1999 Dec; 5 (12) 4021- 4027). HER2/neu mRNA is reported to correlate with the stage of breast cancer (Wasserman L et al., Mol Diagn 1999 Mar; 4 (1) : 21-28). Significant association between the presence of p53 and K-ras gene mutation in the blood and tumor size, depths of invasion, venus invasion and overall survival in patients with colorectal cancer is described in linuma H et al., Int J Cancer 2000 Jul 20; 89 (4) : 337-344; Hardingham JE et al., Mol Med 1995 Nov; 1 (7) : 789-794; and Salbe C et al., Int J Biol Markers 2000 Oct; 15(4): 300-307. The presence of CEA mRNA in the blood is said to be a useful indicator of circulating tumor cells, and to be one of the determinant prognostic factors for patients with colorectal carcinoma (Taniguchi et al., Cancer 2000 Sep 1 ; 89 (5) : 970-976) , and to correlate with a significant shorter survival in patients with non-small cell lung cancer (Yamashita Jl et al., J Thorac Cardiovasc Surg 2000 May; 119(5): 899-905). The simulataneous presence of CEA and CK20 mRNA is said to be a potent prognostic factor (Yamaguchi K et al., Ann Surg 2000 Jul; 232(1 ): 58-65).

The above studies rely upon the hypothesis that the mere presence of circulating cancer cells is indicative of an enhanced risk of a metastasis as the dissemination of cancer cells is known to be a prerequisite in the development of micrometastases and solid metastases. Accordingly, the detection of disseminated cancer cells has been used to predict disease progression and recurrence. While most of said studies use one parameter for detection, some suggest multiple marker RT-PCR analysis in order to provide a more reliable and more sensitive assay than a single molecular marker assay for the detection of disseminated cancer cells. However, these studies fail to assign the markers investigated to disseminated cancer cells. The mere detection of such markers therefore may indicate the presence of circulating cancer cells, but it does not provide reliable information on the molecular endowment of disseminated cancer cells which determines their behaviour and, as a consequence, their relevance for the disease.

WO 99/10528 describes a method for identifying and characterizing disseminated cancer cells with the help of multiple-parameter analyses of cancer-specific and cancer-associated genes. The relevance of such genes to cancer generally arises from a multiplicity of studies which connect clinically diagnosed cancers with one of said genes.

WO 00/06702 describes a method for isolating disseminated cancer cells from body fluids. This method makes it possible to enrich said cancer cells in a gentle manner, in particular in the vital state. This not only allows the possibility of carrying out meaningful genomic DNA analysis and even LOH (LOH = loss of heterocygosity) analysis which requires an at least 1 :1 ratio of cancer cells to non-cancer cells, but also provides disseminated cancer cells for a number of other diagnostic, therapeutic, animal-experimental and scientific tests. Thus, numerous possible applications of disseminated cancer cells are described, for example the identification of novel therapeutic targets; in the context of developing active substances, the screening of active substances, for example identifying and characterizing lead substances and in particular for developing active substances against surface structures of disseminated cancer cells; the choice of a therapy depending on the particular individual and on the stage of the disease.

Apart from specific surgery to remove a primary tumor, measures of cancer therapy are not individually suited to the actual situation of a patient. This is especially true for chemotherapy whose effectiveness is generally very low. Thus it is regular practice nowadays to administer active substances without knowing whether or not the individual concerned will respond to a particular active substance.

Attempts to make the administration of active substances conditional on particular markers expressed by the primary tumor have led to negligible progress in cancer therapy. Besides, this type of approach first requires diagnosing the tumor and taking appropriate material. However, this is usually only possible at an advanced stage of the disease.

It is therefore an object of the present invention to use the earliest possible and at any time repeatable diagnosis and prognosis of a cancer for an appropriately adjusted treatment of the cancer.

This object is achieved according to the invention by the assessment of targets from disseminated cancer cells with respect to their role for the clinical manifestation of cancers.

The present invention therefore relates to a method for the clinical validation of a target from disseminated cancer cells, which method comprises that for a population of individuals it is determined whether a target status determined in disseminated cancer cells of the individuals correlates with at least one cancer-related information about the clinical status of the individuals.

In this connection, the term "target" denotes any biochemical or molecular biological feature of disseminated cancer cells. This includes both therapeutic and diagnostic targets and in particular prognostic targets. Genomic targets at DNA level, for example, are included as well as targets involved in expression, for example those at RNA level, in particular at mRNA level or at protein level. Examples of targets may be mutations, insertions, deletions, LOHs, amplifications, aberrations in the chromosomal set, and the like; expression of splice variants; and also over- and underexpression of particular mRNAs or proteins - and other unusual, in particular cancer-specific modifications of particular cell components.

Preferred targets are those which affect qualitative features of the DNA machinery and/or RNA machinery. Special mention must be made here of targets which affect and, in particular, oncogenically influence cell properties such as cell division, cell growth, cell-cell interactions, inhibition of tumor suppression and therapy resistances and thus have an influence on the clinical picture of a cancer. Targets relating to DNA recombination, DNA amplification, DNA repair, cell cycle inducers and apoptosis inhibitors are in particular included here.

The targets to be validated according to the invention may be targets known per se. Examples which may be mentioned are:

- especially oncogenes and tumor suppressor genes such as p53, genes of the ras family erb-B2, c-myc, mdιm2, c-fos, DPC4, FAP, nm23, RET, WT1, and the like, LOHs, for example with regard to p53, DCC, APC, Rb and the like and also BRCA1 and BRCA2 in hereditary tumors, microsatellite instability of MSH2, MLH1 , WT1 and the like; also tumorous RNAs such as CEA, cytokeratins, e.g. CK20, BCL-2, MUC1 , in particular tumor-specific splice variants hereof, MAGE3, Mud 8, tyrosinase, PSA, PSM, BA46, Mage-1 and the like, or else morphogenic RNAs such as maspin, hCG, GIP, motilin, hTG, SCCA-1 , AR, ER, PR, various hormones and the like;

furthermore, especially RNAs and proteins which affect the etastasizing profile, i.e. the expression of molecules involved in angiogenesis, motility, adhesion and matrix degradation such as bFGF, bFGF-R, VEGF, VEGF-Rs, such as VEGF-R1 or VEGF-R2, E-cadherin, integrins, selectins, MMPs, TIMPs, SF, SF-R and the like, the cell cycle profile or proliferation profile, such as cyclins (e.g. expression ratio of cyclins D, E and B), Ki67, p120, p21 , PCNA and the like, or the apoptosis profile, such as FAS (L+R), TNF (L+R), perforin, granzyme B, BAX, bcl-2, caspase 3 and the like.

These and other targets are described and illustrated in WO 99/10528 and also WO 00/06702. These disclosures are incorporated by reference in their entirety and thus are part of this description.

Moreover, a particular aspect of the present invention relates to the identification of targets from disseminated cancer cells and to the validation thereof. These targets may have a relevance with respect to cancer which is known and which is related to, for example, primary tumor material or which is not established. Thus, identification is aimed, for example, at DNA aberrations such as mutations, amplifications and LOHs. According to a specific embodiment of this aspect, splice variants are identified in disseminated cancer cells.

Splice variants are expression products of particular genes whose mRNA is alternatively spliced. Accordingly, splice variants differ qualitatively, for example with regard to composition and/or sequence of exons and/or by additional incorporation of intron (partial) sequences. According to the previously mentioned modifications, splice variants may occur at mRNA level and in particular may lead to a shift in the reading frame or to the generation of a stop codon. Likewise, they may occur at the protein level and may lead to expression of a protein with modified structure and function, in particular to expression of a truncated protein.

According to a particular embodiment, splice variants are identified by using at least one exon-spanning primer for amplifying a particular mRNA or a transcript thereof, in particular cDNA. Depending on the mRNA structure, two or more exon-spanning primers may be used. In the case of multiple primers, it is expedient to use said primers in a multiplex PCR approach. The amplicons from disseminated cancer cells are compared with amplicons obtained accordingly from normal cells. Normal cells generally mean cells without cancerous modifications. For studying disseminated cancer cells from blood, the preferred normal cells used are leukocytes. The comparison makes it possible to identify amplicons and thus also mRNAs of normal and altered size. The composition of the amplicons and thus of the mRNAs may be determined in a manner known per se, for example by hybridization, in particular on biochips with or without prior amplification, or by sequencing.

This embodiment for identifying splice variants is directed especially at genes having cancer-related function, i.e. in particular genes affecting qualitative features of the DNA machinery and/or RNA machinery. In this connection, mention must be made of, for example, cell-cycle-related genes such as cyclins B, D and E, p21, p53 or the apoptosis-related genes such as caspases 1 to 9, bax, bcl-2, Fas ligand or Fas receptor.

According to another preferred embodiment, splice variants are identified by amplifying mRNA or a transcript thereof, in particular cDNA, by means of degenerate PCR primers specific for conserved, functional domains of particular protein families, for example src domain, SH2 domain or kinase domain. Comparison with accordingly amplified mRNA or a transcript thereof, in particular cDNA, of normal cells leads, according to the previously described procedure when using exon-spanning primers, to the identification of splice variants affecting conserved, functional domains of proteins.

According to another particular embodiment, splice variants are identified by means of differential hybridization known per se, preferably subtraction hybridization (SSH). To this end it is possible to transcribe firstly mRNA or poly(A) RNA of disseminated cancer cells (tester mRNA) and secondly mRNA or poly(A) RNA of normal cells (driver mRNA) into cDNA (tester cDNA and driver cDNA, respectively). It is also possible to transcribe mRNA from disseminated cancer cells into cDNA and to isolate poly(A) RNA from normal cells or to transcribe mRNA from normal cells into sense cRNA. cDNA from disseminated cancer cells and cDNA from normal cells may be compared and hybridized against a cDNA gene library. The gene library may have been generated either from the disseminated cancer cells or from the normal cells. Preference is given to hybridizing tester cDNA from disseminated cancer cells against driver cDNA from normal cells. It is in principle also possible to use poly(A) RNA or sense cRNA instead of the driver cDNA. The hybridization generally uses an excess of cDNA and/or poly(A) RNA or sense cRNA of one of the two cell populations over the other. It is preferred to use an excess of the appropriate nucleic acids from normal cells, i.e. in particular of driver cDNA. According to a particular embodiment, the tester cDNA is initially cut by a restriction enzyme prior to hybridization. The tester cDNA is ligated in two mixtures with two different adapters which represent the binding sites for the primers of a later PCR. Each mixture of the adapter-carrying tester cDNA is hybridized with an excess of driver cDNA.

The non-hybridizing and therefore still single-stranded DNAs or RNAs of interest may be identified and, if required, isolated, and in particular sequenced. Thus, after denaturing and hybridizing the tester cDNA in the presence of an excess of driver cDNA, there are preferably sequences left which are present in the tester cDNA, but not in the driver cDNA. These transcripts can be amplified in a subsequent PCR. To this end, the two adapter mixtures are combined.

It is advantageously possible to carry out a plurality of cycles comprising hybridization and isolating single-stranded molecules remaining in the hybridization mixture. This leads to increasing amplification of those DNAs or RNAs which are substantially present only in one of the two cell populations comparatively studied.

This embodiment especially relates to the identification of splice variants of genes to which no cancer-related function has been assigned yet.

The clinical validation of the invention is based on the assignment of target and clinic. Thus there are on the one hand a particular biochemical or molecular biological feature and on the other hand characteristics of a cancer, in particular , as clinical symptoms.

With respect to a single individual, this implies both determining the status of the target from disseminated cancer cells and providing at least one item of cancer- related information about the clinical status of the individual.

Therefore the present invention also relates to a method for collecting a cancer- related dataset (profile) of an individual, characterized in that a) the individual is examined for disseminated cancer cells, b) optionally at least one aliquot of a sample derived from the body of the individual examined in a) is set aside in a substance library, c) at least one cancer-related information about the clinical status of the individual is provided, and d) the result obtained from the examination in a) and the information provided in c) and, optionally, the sample set aside in b) are linked to each other in the form of a dataset.

The examination of an individual for disseminated cancer cells includes an investigation of whether or not a sample from the body of the individual comprises disseminated cancer cells. Advantageously, this examination of disseminated cancer cells also comprises determination of the status of at least one target from the disseminated cancer cells. For the purpose of clinically validating a target, the status of this target is expediently determined.

This examination is analytical and expediently takes place in vitro. The "status of a target" comprises any appearance of a particular target with respect to quantity and/or quality. Whether determining the status provides a qualitative or quantitative result depends especially on the definition of the target. Qualitative determination of the status is generally directed toward the investigation of whether or not a particular target is present. Determining the status of these targets may be carried out, for example, in such a way that the result is either "positive" or "negative". To this end, it is necessary to define the target in such a way that a particular set of biochemically and/or molecular biologically determinable analytical values is assigned to the results "positive" or "negative". Thus, qualitative determination of the status generally also includes quantitative determination of values, which the definition of the target is based on. It is also possible, for example, to define the overexpression or underexpression of particular mRNAs or proteins as a target in such a way that expressed amounts of a particular mRNA or protein above or below a limit are to be judged "positive" or "negative". According to the invention, qualitative determination of the status is useful, since it simplifies the determination, which is based on statistical evaluation, of whether the target status correlates with the cancer-related information.

Thus, examples may be the detection of particular DNA sequences, in particular of mutated sequences (positive/negative); amounts of particular sequences compared with wild-type sequences or reference sequences or when compared to the amount of the same sequence in non-cancer cells, preferably derived from the same individual and in particular from the same sample, in particular in case of amplifications (e.g. positive, if ratio > 2) and LOHs (e.g. positive, if ratio < 0.5); particular mRNA sequences (positive/negative), in particular splice variants (positive/negative); the presence of particular mRNAs in disseminated cancer cells (positive/negative), amounts of particular mRNAs, in particular reduced or enhanced expressions of mRNAs (positive above a particular amount of mRNA per cell equivalent or when compared to the amount of the same mRNA in non-cancer cells, preferably derived from the same individual and in particular from the same sample); or the presence or absence of a particular protein (positive/negative).

The term cancer cell means according to the invention a cell having one or more modifications connected with cancer, i.e. degeneration in a general sense. This definition is based on the idea that generation of cancer is a continuous modifying process. The path from a normal cell to a cancer cell and in particular to a tumor cell, for example, generally requires a plurality of modifications, in particular of the genetic material or of the expression of the genetic material of cells. The term cancer cell therefore also includes precursors of cancer cells and in particular tumor cells, which have cancerous or tumorous modifications.

The term "disseminated cancer cell" is defined in particular in relation to solid tumors, i.e. especially primary tumors, metastases and recurrences. In contrast to solid tumors, disseminated cancer cells can circulate in the body of an individual. This generally involves transport organs of the body, especially body fluids and in particular blood. Disseminated cancer cells generally derive from a solid tumor by being initially part of a solid tumor, i.e. in particular of tumor tissue, from which they subsequently detach. This causes disseminated cancer cells to leave the area of the body predetermined by the solid tumor, in particular the morphological structural units affected by the tumor, for example the organ, and to reach inter alia locations to which they have no morphological connection based on the solid tumor.

According to a particular aspect, disseminated cancer cells are characterized by their relatively low amount with respect to non-cancer cells equally present. They are therefore also denoted as residual cancer cells (minimal residual disease, or MRD for short). Regarding cell-containing body fluids, for example, the proportion of disseminated cancer cells is generally below 1 :1000, usually below 1 :10,000 and in many cases even below 1:100,000, with respect to the number of non-cancer cells in a randomly obtained sample of body fluid. For blood, these ratios are valid in particular with respect to mononuclear cells (MNC for short).

Generally, disseminated cancer cells are studied in cell-containing mixtures having disseminated cancer cells optionally in addition to non-cancer cells. Depending on the type of analysis to be carried out for determining the target status, the mixtures may have different proportions of disseminated cancer cells. In particular cases, for example LOH analysis or amplification analysis, cancer cell contents of 50% or more may be useful. Accordingly it is preferred to use preparations derived from a body fluid of the individual which have a cancer cell to non-cancer cell ratio of more than 1 :1. Higher ratios of more than 9:1 and in particlar cases of more than 99:1 are even more preferred for determining the status of the target and to assign said status to the cancer cells. The enrichment of disseminated cancer cells, which may be necessary, can be carried out in a manner known per se, for example via immunospecific adsorption methods or, preferably, using size- and/or shape-dependent separation techniques such as the screening technique described in WO 00/06702 which is incorporated herein by reference. The latter method using a screen that retains said cancer cells has the advantage that the disseminated cancer cells are recovered in an essentially unaltered state which reflects the biological state they have in their natural environment, i.e. the body fluid of the individual they have been derived from. Moreover, WO 99/40221 and WO 00/46585 also incorporated herein by reference describe methods for the enrichment of tumor cells by means of density gradient centrifugation using separation mediums having a defined density in the range of 1 ,055 to 1 ,070 g/ml or 1 ,055 to 1 ,065 g/ml, respectively.

In principle, the status of a particular target can be determined in a manner known per se. Choosing suitable methods is within the skilled worker's knowledge and is based in many cases on extraction measures, on amplification reactions of particular nucleic acids, in particular by means of polymerase chain reaction (PCR for short), where appropriate combined with reverse transcription (RT for short), on hybridization experiments, for example from the well-established Southern blot to the use of biochips, and also on immunological detection reactions. For further illustration, reference is made, for example, to WO 99/10528 and WO 00/06702 which are hereby part of the present description.

Preference is given to setting aside in a substance library at least one aliquot of a sample of the body of the individual examined in a). This aliquot may be a part of the sample taken from the body of the individual, in particular of a body fluid such as blood. It is however also possible to use mixtures derived therefrom, in particular cell- containing fractions of body fluids such as serum, buffy coat, the MNC fraction, cell fractions based on disseminated cancer cells, i.e having an increased cancer cell content (preferably a ratio of more than 1 :1 cancer cells to non-cancer cells or even higher), or cell fractions based on normal cells, which are substantially free of disseminated cancer cells, for example leukocyte fractions, and also cell components derived therefrom such as cell extracts, in particular nucleic acid fractions and/or protein fractions. In particular, it is advantageous and in specific cases useful to set aside for comparison both aliquots based on disseminated cancer cells and aliquots based on normal cells. Advantageously, the substance library comprises for a sample taken from an individual at a particular point in time at least two preparations: a first which has been enriched for disseminated cancer cells and a second which is substantially free of disseminated cancer cells.

Providing cancer-related information about the clinical status of an individual relates to his or her history. Included therefore are especially diagnostic assessments which relate to the clinical symptoms of the individual.

This includes, for example, diagnosis of a cancer, in particular of a primary tumor, of a recurrent tumor or of distant metastases, generally including the location of the disease and the time of diagnosis. Preference is given to an appropriate histological result and/or to information about staging, i.e. about the extent of the disease, taking into account, where appropriate, the patient's history and physical examination results, such as family disposition, status of menopause, environmental factors, previous diseases, and symptoms such as tiredness, loss of weight, fever, night sweats, cough, continuous pain, and also modifications of organs such as the skin, lymph nodes, lungs, thorax, abdomen, testicles and prostate, rectum and vagina. Mention must be made in particular of pathological staging with taking of tissue samples, for example mediastinoscopy, bone marrow biopsy, removal of axillary lymph nodes, laparotomy, consideration of particular blood values, for example with respect to blood levels of particular enzymes such as alkaline phosphatase, lactate dehydrogenase or acidic phosphatase, or of the sugar level, and also visualizing examinations, in particular CT and MRI, and ultrasound examinations.

In the context of diagnosis and staging of tumors, the tumor classification represents usually particularly important cancer-related information about the clinical status of an individual. Mention must be made here especially of TNM classification which is known per se and classifies tumors according to their anatomic extent. One aspect of this system relates to the assessment of this primary tumor itself. Of particular importance according to the invention are the statements "no indication of a primary tumor (TO)" and "size of the primary tumor (T1-T4)". Another aspect relates to the assessment of regional lymph nodes. Of particular importance here are the statements "no regional lymph node metastases (NO)" and "metastatic involvement of regional lymph nodes (N1-N3)". Another aspect relates to the assessment of distant metastases, the statements "no distant metastases (MO)" and "distant metastases, stating the attacked organ(s), where appropriate (M1)" being of particular importance here. Yet another aspect relates to the assessment of residual tumors (local recurrences). Here, the statements "no residual tumour (RO)", "microscopic residual tumour (R1 )" and "macroscopic residual tumor (R2)" are particularly important according to the invention. Another aspect of staging relates to the assessment of the histopathological degree of differentiation (grading), the statements "well differentiated (G1)", "moderately differentiated (G2)", "badly differentiated (G3)" to "undifferentiated (G4)" being particularly important according to the invention.

A type of cancer-related information which is preferred according to the invention is determined by the moment when the individual dies. This moment may be related to a first point in time, for example to the moment when a cancer, for example a primary tumor, a local recurrence or a distant metastasis, is diagnosed, to the moment when disseminated cancer cells are taken, in particular when the target status is determined, or when a therapeutic treatment commences or ends. The information is generally given in form of a period which is to be denoted according to the invention by the term "survival time". The period relates to the first, in principle deliberately selectable point in time, preferably the moment of making the diagnosis or, in particular, of taking disseminated cancer cells from the individual. In the case of a plurality of individuals, the information is given as an average. As an alternative to survival time, it is also possible to give the survival rate, i.e. the percentage of individuals surviving a particular period.

Another type of cancer-related information which is preferred according to the invention is the so-called "cancer-free, i.e. tumor-free and in particular recurrence- free" period. On the one hand this period is determined by a first point in time, for example the time of making the diagnosis, the moment when a cancer of an individual is, according to the criteria above, not diagnosable (any longer), for example after removing the tumor, and/or the moment of taking disseminated cancer cells. On the other hand this period is determined by a second point in time when in the same individual a cancer, i.e. in particular a renewed tumor (recurrence), where appropriate as local recurrence or as distant metastasis, is or is not diagnosed. The first point in time is in principle deliberately selectable under the above conditions. It is expediently determined by the moment of removing the tumor or, in particular, of the taking of disseminated cancer cells. The second point in time more or less defines the clinical endpoint and indicates the clinical importance of cancer-related information. Examples of suitable statements are "no tumor", "tumor after less than 2 years" and "tumor after 2-5 years".

A particular item of cancer-related information is the qualitative and, optionally appropriate, also quantitative assessment of disseminated cancer cells of an individual. Of particular importance in this connection is the status of one or more than one target (geno- and phenotyp) of the disseminated cancer cells which, once clinically validated according to the invention, reflects the item(s) of cancer-related information it correlates with. Also of particular importance in this connection are a comparison in this regard at various times and, based thereupon, the finding of whether the amount and/or type of disseminated cancer cells have changed, in particular increased or decreased, generally with respect to one or more parameters, in particular targets. This information especially reflects the progress of the cancer and can provide direct evidence for taget-related therapeutic effectiveness.

A further particular aspect of the present invention relates to the indication of therapeutic measures as cancer-related information. Examples of suitable information are "no therapy", in particular "no surgery", "no systemic therapy"; "therapy", in particular "surgery" or "systemic therapy", e.g. "chemotherapy, optionally indicating the active substance(s) and/or therapy plan", "radiation (radiotherapy)", "adjuvant therapy", for example "hormone therapy", "immunomodulation" or "target-related therapy", or combinations thereof.

Preferably the above steps c) and d) are repeated for the same individual at another time, i.e. the individual is again clinically examined and the cancer-related information obtained is added to the already existing dataset (follow-up). Likewise steps a) and optionally b) and also d) may be repeated for the same individual at another time (longitudinal assessment) so that a plurality of samples from an individual, which have been taken and optionally set aside at different times, can be added to the dataset. Thus, the substance library, according to a particular aspect of the invention, includes samples which have disseminated cancer cells or cell components derived therefrom and which were taken from the same individual at various times. It is then advantageously possible to assign cancer-related information about the clinical status of an individual to samples, which have been taken at different times, for example based on the moment of taking the samples. Moreover, it is advantageous to provide the possibility to facilitate a time-based comparison.

To clinically validate a particular target, the target status which is determinable in disseminated cancer cells of an individual is assigned to at least one item of cancer- related information about the clinical status of the individual, where appropriate with respect to the moment when the disseminated cancer cells were taken. This assignment is carried out for a plurality of individuals.

In the case of a multiplicity of individuals, it is possible and expedient to store cancer- related datasets of single individuals which are to be used according to the invention in the form of databases. This makes it possible to assign individual items of information firstly to an individual and the cancer-related information linked thereto about the clinical status and secondly to particular disseminated cancer cells and their target-related status. Accordingly, the database may be defined to comprise datasets which assign one or more than one item of clinical information to one or more than one target status.

Databases in accordance with the invention are collections or compilations of data, numbers, facts or other elements which may be compiled, stored and accessed using electronic, electromagnetic, electrooptical or similar methods. Said elements are arranged systematically or methodically and are individually accessible (structure). The database also includes those elements which are necessary for running or using the database, for example thesaurus or indexing systems.

The present invention therefore also relates to such databases, the building and use thereof for clinical validation of targets from disseminated cancer cells. The data comprise at least one item of cancer-related information for each of various individuals and the assignment of this information to disseminated cancer cells of each individual. Preferably, the data also comprise the status of at least one target from the disseminated cancer cells. Usually, it comprises the status of more than one target from disseminated cancer cells. Expediently, the database makes it possible to group individuals on the basis of one or more item of cancer-related information and, where appropriate, also on the basis of the status of one or more targets from the disseminated cancer cells. Setting up the database includes adding cancer-related information about individuals already recorded in the database and also adding further individuals. Adding information of further individuals leads to the corresponding addition of disseminated cancer cells from said individuals or, where appropriate, of information about the status of one or more targets from the disseminated cancer cells. If further cancer-related information about individuals already recorded in the database is added, then it may be useful to add further disseminated cancer cells or, where appropriate, information about the status of one or more targets from said disseminated cancer cells.

Generally the individuals to be included in the validation are selected using at least one item of cancer-related information and are combined in a group (population). A population is thus a group of individuals having at least one cancer-related common feature. Such a group is formed, for example, by individuals having a particular history such as mastophathy or a diagnosed and possibly surgically removed tumor of a particular type, for example patients having had breast carcinoma surgery. The status of the target from corresponding disseminated cancer cells is assigned to each of the individuals of this population. This assignment makes it possible to deduce the proportion of individuals of this population, who have a particular target status, compared with the total number of individuals in this population. Whether this target status correlates with at least one further item of cancer-related information about the clinical status of the individuals is based on a comparison. For this purpose, a subpopulation is formed among the individuals included in the population, using said further cancer-related information about the clinical status. The status of the target from corresponding disseminated cancer cells is again assigned to this subpopulation. This assignment makes it possible to deduce the proportion of individuals of a specific population who have the target status. Comparison of the proportion of individuals of the population firstly with the proportion of individuals of the subpopulation and secondly with respect to a particular target status forms the basis for the correlation analysis.

It is in particular possible to combine a plurality of items of cancer-related information and this leads to the formation of populations or subpopulations defined accordingly. This is particularly relevant regarding therapeutic measures in connection with the above-described cancer-related information, especially with the survival time and the cancer-free period.

According to one embodiment of the present invention, the individuals may be divided up into the following groups, depending on the time of determination of the staus: patients with precancerous lesions; patients having a manifest tumor (which may have been treated, e.g. surgically removed, or not) prior to systemic therapy; patients after systemic therapy, in particular with a short-term observation of from about 3 to 24, preferably 5 to 12 and in particular 6 to 10 months, a medium-term observation of from 2 to 5 years or a long-term observation of more than 5 years after systemic therapy. Individuals with recurrence can be included as a further group which in turn can be furher devided up into individuals before systemic therapy and individuals after systemic therapy.

According to another embodiment of the present invention, these groups of individuals may then be divided up using further cancer-related information, for example the survival time and preferably the cancer-free period. Accordingly it can be determined whether or not survival time or cancer-free period correlates with the status of a target in one of these groups. This may also provide information on the choice of therapy if, for instance, survival time or cancer-free period correlates with status of the target for a population of individuals having undergone a particular therapy while there is no correlation in other populations of individuals having undergone a different therapy or no therapy. Thus, according to a particular embodiment, the invention relates to a method for clinically validating the therapeutic relevance of a target from disseminated cancer cells, which comprises determining whether a status of said target correlates with at least one item of cancer-related information about the clinical status of at least one population of individuals who have undergone a first systemic therapy and at least for one further population of individuals who have undergone no or a different systemic therapy. Said item of cancer-related information preferably is selected from survival time, cancer-free period or changes of the amount of disseminated cancer cells having said target status.

According to a particular embodiment, the above groups of individuals are divided up into at least 3 subgroups having a particular target status: (a) no tumor (primary or recurrence, respectively) within the observation period (b) early tumor (primary or recurrence, respectively), e.g. within the first half of the observation period, in particular after less than 2 years from the target status determination (c) late tumor (primary or recurrence, respectively), e.g. within the second half of the observation period, in particular after 2-5 years.

The observation period advantageously starts when the individuals are examined for disseminated cancer cells, i.e. the status of the target to be evaluated is determined. Moreover, it is also possible to select other points in time, e.g. the time of diagnosis of the tumor or its removal by surgery.

This allows to provide molecular algorithms, i.e. a target status or combinations of more than one target status, which are, according to one of said subgroups, prognostically associated with one of the above groups of individuals.

Disseminated cancer cells characterized by such a molecular algorithm can be directly used to validate further targets from disseminated cancer cells. Since said algorithms reflect at least one item of cancer-related information about the clinical status of the corrsponding individual, a correlation between the status of a particular target and one of said algorithms indicates a correlation between the status of said target and said cancer-related information.

The evaluation by correlation analysis is carried out using statistics, in particular by means of suitable bio-mathematical methods, for example by logistic regression (cf. J.

Hϋsler, H. Zimmermann: Statistische Prinzipien fur medizinische Projekte [Statistical principles for medical projects], Verlag Hans Huber, 2nd, expanded edition, 1996 Bern, Switzerland; Glenn A. Walker: Common statistical methods for clinical research, SAS Institute, 1997 Cary, NC, USA). In this context, the influence of various factors on a target parameter having a categorical or ordinal scale of values is analyzed. Therefore, usually a sufficiently large population or sufficiently large subpopulations of individuals are used. The size of the population(s) depends on the desired statistical significance of the statistical analysis to be undertaken according to the invention.

According to the invention, the target is clinically validated in the case of a statistical significance of p < 0.10 with 90% significance, preferably of p < 0.05 with 95% significance and in particular of p < 0.01 with 99% significance. This clinical validation relates in a narrower sense to the connection between the presence in an individual of disseminated cancer cells characterized according to the target status, and the individual's clinical status represented by the cancer-related information included in the evaluation. However, in a broader sense this also means a connection between the target as a feature of disseminated cancer cells and the clinical picture of a cancer.

Thus the clinical validation according to the invention of a target includes in particular the finding of whether the evaluated target status leads to progression of the cancer (driver) or protects against it (protector). This finding is of principal importance for diagnosis and prognosis, and especially for a target-specific cancer therapy.

The distinction between driver targets and protector targets may be made, for example, using survival time analysis. The survival function describing the survival time indicates the probability of surviving until a particular point in time. The statistical estimation of the survival function may be carried out, for example, using the Kaplan- Meier method (also called product limit). The survival functions firstly of individuals having a particular status of the target to be evaluated and secondly of individuals not corresponding to this status are compared using the log rank test or the Wilcoxon test. Here too, it is necessary for two survival functions to be different in the case of a statistical significance of p < 0J0 with 90% significance, preferably of p < 0.05 with 95% significance and in particular of p < 0.01 with 99% significance.

In suitable cases and in view of a treatment carried out on an individual, it is possible for the clinical validation of the invention to facilitate distinguishing between those individuals who respond to the treatment (responders) and those who do not (non- responders). This finding too is of fundamental importance for the choice of therapy and in the end ensures increased effectiveness on a particular therapeutic measure, e.g. on administration of particular active substances. The purpose of the method of the invention is the clinical validation of targets from disseminated cancer cells. If the determination according to the method results in a correlation between the target status and cancer-related information, then the target is clinically validated according to the invention. Depending on the correlating cancer- related information, the clinical validation comprises a connection of particular aspects of the clinical picture of cancers with the target status. In particular, the dependence on defined clinical endpoints such as the death of the individual and/or the appearance of recurrences is determined (validation as an prognostic indicator), where appropriate taking into account particular therapeutic measures, in particular a systemic therapy (validation as a therapeutic indicator). In particular, the present invention aims at the clinical validation of targets from disseminated cancer cells as independent indicators, i.e. where the target status does not correlate with primary tumor-related information, such as the TNM-claasification.

According to a particular embodiment, combinations of a plurality of targets are clinically validated. The finding that the status of a first target in combination with the status of one or more further targets correlates with at least one item of cancer- related information about the clinical status of an individual is carried out in analogy to the above-described procedure. Instead of a particular status of a target, a particular combination of statuses with respect to a plurality of targets is assigned to the populations or subpopulations. This results in molecular algorithms which can be used to validate further targets, even without having recourse to the individuals clinic status, simply be determining whether the status of the target to be evaluated correlates with a particular molecular algorithm.

Advantageously, the clinical validation is carried out by comparison with controls which have no disseminated cancer cells carrying the target to be validated.

Targets from disseminated cancer cells, which have been clinically validated according to the invention, then form the basis for target-related dealing with cancers. This includes an appropriate diagnosis and a target-related therapy.

An important aspect of a target-related therapy concerns judging whether a functional relationship is present between a particular active substance and a clinically validated target. Thus, the present invention further relates to a method for the functional validation of the targets, preferably of targets clinically validated according to the invention. For this purpose, the active substance is allowed to act ex vivo on disseminated cancer cells whose reaction is then determined. It is, for example, possible to choose at least one measurement parameter whose modification is determined in the disseminated cancer cells. The disseminated cancer cells expediently correspond to a particular target status; in particular they possess this target (positive). As a control, the measures of the method may be carried out accordingly on cells not corresponding to the target status, i.e. in particular cells not possessing the target (negative). It is generally possible to have recourse to cytobiological test systems known per se. If necessary, disseminated cancer cells can be kept in culture and suitable bioassays may be carried out. In this way it is for example possible to use active substances known per se and having anti-neoplastic action and/or to use active substances employed for adjuvant therapy. In particular it is possible to test target-related active substances.

According to another aspect, the above method also allows testing active substances on targets from disseminated cancer cells and functionally validating them. This applies in particular to target-related active substances. This is an important aspect of active substance development according to which potential active substances can be selected in a target-related manner - for example in the context of screening methods - and subsequently be validated.

The purpose of this method for the functional validation of target and/or active substance is detecting active-substance-dependent, molecular and/or morphological modifications in the disseminated cancer cells. If the determination of one or more parameters results in a state in the disseminated cancer cells which, after the action of the active substance, is different from the state prior to the action of the active substance, then the target is functionally validated with respect to the active substance, or the active substance is functionally validated with respect to the target, according to the invention. The functional validation comprises in particular a functional connection between active substance and target in disseminated cancer cells.

It is possible, where appropriate, for the present functional validation in disseminated cancer cells to build on a functional prevalidation of the target in other cell systems. For example, targets may be cloned and expressed in a manner known per se. To this end, suitable cell systems, in particular human cell lines, which can be transfected or transformed appropriately, are available to the skilled worker. Such target-carrying cell systems may be contacted with one or more active substances in the manner already described above. This method too serves to establish a molecular and/or morphological algorithm of action, which in turn can be validated according to the invention in disseminated cancer cells.

Depending on the type of target, active substances having antisense function, antibody function or ribozyme function are preferred. This type of active substances is advantageous according to the invention, since they can be specifically developed and optimized with respect to a target clinically validated according to the invention.

According to another aspect, active substances having antisense function, antibody function or ribozyme function may also serve to characterize biological control systems, in particular biological regulatory circuits. This in turn may lead to the identification of a drug target which is different from the actual target, which has been clinically and functionally validated with respect to the antisense molecule, immunoglobulin or ribozyme used, but interacts with the latter via the characteristic control system.

For example, progesterone receptor expression in disseminated cancer cells is clinically validated according to the invention. Estrogen receptor expression is dependent on progesterone receptor expression. Both are regulated via a shared control system. The estrogen receptor is a suitable drug target which may be treated, for example, with tamoxifen.

To characterize appropriate biological control systems it is possible to use, for example, antisense molecules for the knockout or knockdown of particular mRNA targets. It is possible to infer from this the participation and, where appropriate, function in a biological control system via particular measurement parameters. The biological control system may be characterized in established tumor cell lines, normal cells or, preferably, in expediently isolated disseminated tumor cells. Methods for characterizing a biological control system are known per se, for example knockout or knockdown of a parameter using antisense molecules, biochemical analysis, determination of gene expression using well-established methods or biochips, and bioinformatics. Characterization of the biological control system may identify upstream or downstream drug targets which, owing to their structure (e.g. enzymes, receptors), serve as drug targets for developing, for example, specific enzyme inhibitors or receptor blockers.

In particular, targets on the cancer cell surface may be accessible to both antisense approaches and antibody approaches, while antisense molecules in particular are available for intracellular targets.

Useful antisense molecules are characterized by their specific binding to the target RNA. Expediently, at least parts of the antisense sequence are complementary to partial sequences of the RNA. Preference is given to using nucleic acids which are not cleaved by particular cellular enzymes, in particular RNases. Furthermore, a certain rigidity is advantageous. Thus LNAs (locked nucleic acids) are particularly preferred.

Useful active substances having antibody function are in particular immunoglobulins, for example antibodies or functional fragments thereof, in particular human, animal, polyclonal, monoclonal and recombinant immunoglobulins, antibody fragments, for example Fab, Fab', F(ab)₂, synthetic immunoglobulins. They are in particular characterized by binding to protein targets. This binding advantageously takes place in such a way that the cancer-related relevance of the target is reduced. A typical example is Herceptin.

The clinical and functional validation of a target from disseminated cancer cells, according to the present invention, provides an advantageous basis for developing and testing target-related active substances. The focus on targets clinically and functionally validated uniformly in disseminated cancer cells allows active substance development including pharmacogenomic and toxicogenomic aspects to reduce unwanted side effects and including a correct stratification of patients, i.e. an, if necessary time-dependent, individualized application of active substances. Compared to conventional active substance developments, substantial savings in costs and time are obtained.

Accordingly, the targets clinically and also, where appropriate, functionally validated according to the invention offer the advantage to have the possibility of an individualized therapy. Thus, the present invention also relates to a method for cancer treatment, where the status of at least one validated target from disseminated cancer cells of an individual is determined and a therapy is chosen depending on the status.

The determination of the status represents a diagnostic measure and relates in particular to the characterization of disseminated cancer cells from body fluids, in particular blood. The characterization includes both the identification and detection of the cancer cells as such and the determination of one or more parameters in these cancer cells. In view of human individuals or non-human, animal individuals, this procedure relates in particular to the method for characterizing disseminated and micrometastasized cancer cells using DNA and/or RNA as described in WO 99/10528.

Part of the treatment method of the invention is therefore also a method for characterizing disseminated cancer cells using DNA, RNA and/or proteins, in which the cancer cells which, if required, have been enriched from the body fluid of an individual are tested for at least one clinically and preferably also functionally validated target. It is advantageous and in particular cases useful to carry out the same test on non-cancer cells of the same individual for comparison. This method does not exclude further diagnostic measures. Particular aspects and embodiments of this method result with respect to the methods disclosed in claims 1 to 10 of WO 99/10528, in particular regarding the genes and proteins mentioned in the glossary and further combinations of particular genes disclosed in WO 99/10528, for undertaking multi-parameter expression analyses which are preferred according to the invention and genomic tests for oncogenes and/or mutated tumor suppressor genes.

The choice of therapy falls into the expanded field of diagnosis, too. It allows choosing from particular therapies an optimal therapy plan which includes the type of active substance(s), dosage and administration schedule, while weighing up the benefit/risk ratio. The choice of therapy thus includes in particular the choice between radiotherapy, chemotherapy and adjuvant therapy, the choice of the active substance class or of particular active substances, for example antiandrogens, antiestrogens, aromatase inhibitors, Herceptin, Panorex, antifolates, 5-FU/FA, 6-thiopurines, taxanes, cisplatin and analogs, anthracyclines, metalloprotease inhibitors, angiogenesis inhibitors, differentiation-inducing active substances, nonspecific immunomodulators, and the like, and the choice of target-related active substances. Furthermore, the choice of therapy includes a choice of dosage depending on the disseminated cancer cells, in particular depending on the amount and/or the status of disseminated cancer cells of the individual to be treated.

Moreover, a particular aspect of this method relates to the possibility of therapy monitoring according to which the status of at least one validated target is repeatedly determined at different times and the therapy already selected is assessed and, if necessary, adjusted depending on the state and in particular depending on changes in the status. A method of this type expediently includes the status of a plurality of targets, too.

The treatment per se, i.e. carrying out one or more therapeutic measures, is directed according to the invention at disseminated cancer cells and the indications linked thereto and is generally based on the use of at least one therapeutic agent, in particular one active substance in an effective amount for administration, or on the use of this therapeutic agent, in particular active substance, for preparing an appropriate agent. Appropriate agents, usually in the form of pharmaceutical compositions comprising said active substance and optionally further acitve substances and/or pharmaceutically ingredients for suitable formulation , such as carriers, diluents etc., are well known in the art.

The user of this method is expediently provided with suitable agents for carrying out this method of treatment, for example in the form of packed units (kits). Accordingly, the present invention therefore relates to kits for target-related cancer therapy, which comprise - agents for determining the status of at least one validated target in disseminated cancer cells of an individual; - instructions for the therapy taking account of the determined status.

In addition to these kits, the user may have recourse to the appropriate therapeutic agent, generally an active-substance-containing pharmaceutical.

Accordingly, kits of the invention relate to the combination of firstly a diagnostic component related to disseminated cancer cells and secondly a therapeutic component.

When comparing cells from a primary tumor or from metastases with disseminated cancer cells, the latter are usually different with respect to the status of targets validated according to the invention. In this respect disseminated cancer cells are a tumor entity independent of the primary tumor or metastases and form, according to the invention, the basis for a more sophisticated and more efficient cancer therapy. A particular advantage is gained from clinically validating a target even before clinical testing in the conventional sense, in particular before phase II. Another advantage is the possibility of being able to test active substances on humans ex vivo, and this is of great importance, in particular for identifying lead substances and especially for the validation thereof prior to the complicated preclinical phase of pharmacology and toxicology. Thus, on the basis of the present invention, a considerably shorter time for for developing active substances for more effective treatment of cancer is ensured than is currently the case.

According to a particular embodiment the present invention relates to a method for clinically validating the expression of cytokeratin 20 (target). This validation relates especially to carcinomas. The determination of the status of this target is based on the positive or negative detection of cytokeratin-20 mRNA in cell-containing body fluids, in particular blood, or in fractions derived therefrom, in particular the MNC fraction, for example by means of RT-PCR. The individuals are generally patients having carcinomas which have been diagnosed and, where appropriate, already surgically treated, in particular patients having breast carcinomas. Information about the primary tumor, therapy measures and recurrence status of a plurality of individuals is provided. The evaluation of a sufficiently large population of individuals shows no significant correlation with conventional, histological classification parameters, such as lymph node status, grading or tumor size. In contrast, the positive detection of cytokeratin-20 mRNA correlates with the early appearence of recurrences. Furthermore, and with respect to recurrence-free individuals, negative cytokeratin-20-mRNA expression correlates with chemotherapeutic and radiotherapeutic measures, in particular with anthracycline (AC) therapy.

Thus, the result of clinically validating cytokeratin-20 mRNA from disseminated cancer cells, carried out according to the invention, is that cytokeratin 20-expressing disseminated tumor cells are accessible to a combination of chemotherapy and radiotherapy. In particular chemotherapy according to the AC plan has a cytoreductive effect. A corresponding method for clinically validating CK19 mRNA results, with respect to the same individuals, in no significant correlation between CK19 mRNA expression of disseminated cancer cells and chemotherapy, radiotherapy or hormone therapy or combinations thereof.

The present clinical validation of cytokeratin-20 mRNA opens up the possibility of providing target-related active substances, in particular antisense molecules complementary to cytokeratin-20 mRNA; of testing active substances on cytokeratin 20 mRNA-expressing disseminated cancer cells and determining ex vivo the effect of a particular active substance on cytokeratin 20 mRNA expression; moreover, of treating carcinomas, in particular gynecological carcinomas such as breast carcinomas and ovarian carcinomas by determining cytokeratin-20 expression in disseminated cancer cells of an individual and, in the case of a positive result, choosing a combination of chemotherapy and radiotherapy, with chemotherapy preferably carried out according to the AC plan.

According to another particular embodiment the present invention relates to a method for clinically validating expression of the progesterone receptor (target). This validation relates especially to carcinomas. The determination of the status of this target is based on the positive or negative detection of PR mRNA in cell-containing body fluids, in particular blood, or in fractions derived therefrom, in particular the MNC fraction, for example by means of RT-PCR. The individuals are generally patients having carcinomas which have been diagnosed and, where appropriate, already surgically treated, in particular patients having breast carcinomas. Information about the primary tumor, therapy measures and recurrence status of a plurality of individuals is provided. The evaluation of a sufficiently large population of individuals shows no significant correlation with conventional, histological classification parameters, such as lymph node status, grading or tumor size, but does so with the menopausal status, in case a hormone therapy had been carried out. Furthermore, and with respect to recurrence-free individuals, positive detection of PR mRNA correlates with a reduced number of recurrences and with hormone therapy, in particular tamoxifen therapy, while negative PR mRNA expression correlates with chemotherapeutic and radiotherapeutic measures. The probability of recurrences is substantially lower for positive detection of PR mRNA than for negative detection of PR mRNA. Thus, the result of clinically validating PR mRNA from disseminated cancer cells, carried out according to the invention, is that PR-expressing disseminated tumor cells are accessible to a hormone therapy, in particular a therapy using tamoxifen. This provides the possibility of identifying those patients who respond to tamoxifen (responders).

The present clinical validation of PR mRNA opens up the possibility of providing target-related active substances, in particular antisense molecules complementary to PR mRNA; of testing active substances on PR mRNA-expressing disseminated cancer cells and determining ex vivo the effect of a particular active substance on PR mRNA expression; moreover, of treating carcinomas, in particular gynecological carcinomas such as breast carcinomas and ovarian carcinomas by determining PR expression in disseminated cancer cells of an individual and, in the case of a positive result, choosing a hormone therapy. Moreover, a diagnostic method is provided which comprises dertermining PR mRNA expression in disseminated cancer cells of an individual wherein a positive result indicates a good prognosis, in particular a low risk for recurrences and a prolonged survival time.

According to another particular embodiment the present invention relates to a method for clinically validating overexpression of bcl-2 (target). This validation relates especially to carcinomas. The determination of the status of this target is based on the detection of bcl-2 mRNA. The status is positive if, in a cell fraction that is enriched for disseminated cancer cells obtained from cell-containing body fluids, in particular blood, or from fractions derived therefrom, in particular the MNC fraction, more bcl-2 mRNA is detected - for example by means of RT-PCR - than in a cell fraction tested for comparison which has a relatively low cancer cell content and thus is, in particular, substantially free of disseminated cancer cells. The individuals are generally patients having carcinomas which have been diagnosed and, where appropriate, already surgically treated, in particular patients having breast carcinomas. Information about the primary tumor, therapy measures and recurrence status of a plurality of individuals is provided. The evaluation of a sufficiently large population of individuals shows no significant correlation with conventional, histological classification parameters, such as lymph node status, grading or tumor size. However, overexpression of bcl-2 mRNA correlates with an increased number of recurrences and with chemotherapy and hormone therapy, in particular tamoxifen therapy. The probability of recurrences is substantially higher for detection of increased bcl-2- mRNA expression than for negative detection or in comparison with the detection of normal expression in benign control cells from the patients.

The inventive clinical validation of bcl-2 mRNA opens up the possibility of providing target-related active substances, in particular antisense molecules complementary to bcl-2 mRNA; of testing active substances on bcl-2 mRNA-overexpressing disseminated cancer cells and determining ex vivo the effect of a particular active substance on bcl-2 mRNA expression; moreover, of treating carcinomas by determining bcl-2 expression in disseminated cancer cells of an individual and, in the case of a positive result, choosing radiotherapy rather than chemotherapy or hormone therapy, in particular rather than tamoxifen therapy. Moreover, a diagnostic method is provided which comprises dertermining bcl-2 overexpression in disseminated cancer cells of an individual wherein a positive result indicates a poor prognosis, in particular a high risk for recurrences and a reduced survival time.

According to another particular embodiment the present invention relates to a method for clinically validating expression of ErbB2, the combined expression of CyclinB and -D, the combined expression of ErbB2, CyclinB and -D, and the combined expression of ErbB2, CyclinB, -D and bcl-2 (targets). This validation relates especially to carcinomas. The status is positive if, in a cell fraction that is enriched for disseminated cancer cells obtained from cell-containing body fluids, in particular blood, or from fractions derived therefrom, in particular the MNC fraction, more mRNA is detected - for example by means of RT-PCR - than in a cell fraction tested for comparison which has a relatively low cancer cell content and thus is, in particular, substantially free of disseminated cancer cells. The individuals are generally patients having carcinomas which have been diagnosed and, where appropriate, already surgically treated, in particular patients having breast carcinomas. Information about the primary tumor, therapy measures and recurrence status of a plurality of individuals is provided. The evaluation of a sufficiently large population of individuals shows no significant correlation of ErbB2 or combined CyclinB/D mRNA overexpression with conventional, histological classification parameters, such as lymph node status, grading tumor size or menopausal status. Furthermore, and with respect to recurrence-free individuals, overexpression of ErbB2 or combined CyclinB/D mRNA does not correlate with a reduced number of recurrences. However, the probability of recurrences is substantially higher for combined overexpression of ErbB2, CyclinB and -D, and for combined overexpression of ErbB2, CyclinB, -D and bcl-2 than for normal expression of said mRNAs. Furthermore, and with respect to recurrence-free individuals, negative ErbB2 and bcl- 2-mRNA overexpression correlates with adjuvant therapeutic measures, in particular with Herceptin therapy.

Thus, the result of clinically validating ErbB2, CyclinB, -D and/or bcl-2 mRNA from disseminated cancer cells, carried out according to the invention, is that ErbB2- and bcl-2-overexpressing disseminated tumor cells are accessible to antibody therapy. In particular adjuvant therapy with herceptin has a cytoreductive effect.

The inventive clinical validation of ErbB2, CyclinB, -D and/or bcl-2 mRNA opens up the possibility of providing further target-related active substances, in particular antisense molecules complementary to said mRNAs; of testing active substances on ErbB2, CyclinB, -D and/or bcl-2 mRNA-expressing disseminated cancer cells and determining ex vivo the effect of a particular active substance on ErbB2, CyclinB, -D and/or bcl-2 mRNA expression; moreover, of treating carcinomas, in particular gynecological carcinomas such as breast carcinomas and ovarian carcinomas by determining ErbB2, CyclinB, -D and/or bcl-2 overexpression in disseminated cancer cells of an individual and, in the case of a positive result, choosing an adjuvant therapy, preferably with a target-related active substance such as Herceptin. Moreover, a diagnostic method is provided which comprises dertermining CyclinB, -D and/or bcl-2 overexpression in disseminated cancer cells of an individual wherein a positive result indicates a poor prognosis, in particular a high risk for recurrences and a reduced survival time.

In the drawings,

Figure 1 shows the dependence of the proportion of those patients without recurrence on the recurrence-free period in months for patients having positive PR mRNA detection ( ) and patients having negative PR mRNA detection

(-);

Figure 2 shows the dependence of the proportion of those patients without recurrence on the recurrence-free period in months for patients having bcl-2 mRNA overexpression ( ) and patients having normal bcl-2 mRNA expression ( — );

Figure 3 shows the dependence of the proportion of those patients without recurrence on the recurrence-free period in months for patients having ErbB2 mRNA overexpression ( ) and patients having normal ErbB2 mRNA expression (-

---);

Figure 4 shows the dependence of the proportion of those patients without recurrence on the recurrence-free period in months for patients having CyclinB and CyclinD mRNA overexpression ( ) and patients having normal

CyclinB and CyclinD mRNA expression ( );

Figure 5 shows the dependence of the proportion of those patients without recurrence on the recurrence-free period in months for patients having combined overexpression of ErbB2, CyclinB and CyclinD mRNA ( ) and patients having normal ErbB2 and CyclinB and CyclinD mRNA expression ( );

Figure 6 shows the dependence of the proportion of those patients without recurrence on the recurrence-free period in months for patients having combined overexpression of ErbB2, bcl-2, CyclinB and/or CyclinD mRNA ( ) and patients having normal expression of these parameters ( );

The following examples are intended to serve to illustrate the present invention.

Example 1 : Cytokeratin-20 mRNA (CK20 mRNA)

(a) Determination of the status of CK20 mRNA from disseminated cancer cells

5 ml of peripheral blood was taken in each case from 684 patients having surgically treated breast cancer. Mononuclear cells (MNC) were isolated by means of density gradient centrifugation. If present, disseminated cancer cells are in this cell fraction, too. RNA is isolated from the MNC fraction in a manner known per se and transcribed into cDNA by means of random hexamers and reverse transcriptase. CK20 mRNA- specific detection is then carried out by means of PCR and using the 5'-nuclease detection system known per se (TaqMan®). The following primer sequences and probe sequences were used: SEQ ID NO:1-3. As a control, blood from healthy individuals was analyzed accordingly.

(b) Cancer-related information about the clinical status of the individuals

Where present, information about the lymph node status, about grading, about tumor size, about the menopausal status, about the recurrence status and about therapeutic measures previously carried out of the patients examined in (a) was compiled.

(c) Statistical evaluation by means of logistic regression

In 29% of the patients examined, CK20 mRNA was detected in the MNC fraction (Table 1 ). The examination of healthy individuals resulted in negative detection of CK20 mRNA so that a specificity of 100% can be assumed. There is no correlation of lymph node status (p = 0.81 for n = 335), grading (p = 0.88 for n = 335), tumor size (p = 0.48 for n = 335) and menopausal status (p = 0.80 for n = 335) with CK20 mRNA expression (Table 2). This is in particular true for those patients who had had no chemotherapy and had been recurrence-free for at least two years since diagnosis (Table 3; lymph node status (p = 0.591 for n = 67), grading (p = 0.45 for n = 67), tumor size (p = 0.40 for n = 67) and menopausal status (p = 0J6 for n = 67)). In contrast, CK20 mRNA detection correlates with the early appearance of recurrences (Table 4; recurrence-free versus recurrence < 2 years: p = 0.04 for n = 256). The combination of chemotherapy and radiation correlates with a reduction of CK20 mRNA (Table 5). With respect to patients having had only chemotherapy, the AC therapy correlates with negative detection of CK20 mRNA (Table 6; CMF v. AC p=0.18 for n = 55; no chemotherapy v. AC p=0.20 for n = 109).

Example 2: Progesterone receptor mRNA (PR mRNA)

(a) Determination of the status of PR mRNA from disseminated cancer ceils

5 ml of peripheral blood was taken in each case from 795 patients having surgically treated breast cancer. Mononuclear cells (MNC) were isolated by means of density gradient centrifugation. If present, disseminated cancer cells are in this cell fraction, too. RNA is isolated from the MNC fraction in a manner known per se and transcribed into cDNA by means of random hexamers and reverse transcriptase. PR-specific detection is then carried out by means of PCR and using the 5'-nuclease detection system known per se (TaqMan®). The following primer sequences and probe sequences were used: SEQ ID NO:7-9.

As a control, blood from healthy individuals was analyzed accordingly.

(b) Cancer-related information about the clinical status of the individuals

Where present, information was compiled about the lymph node status, about grading, about tumor size, about the menopausal status, about the recurrence status and about therapeutic measures previously carried out.

(c) Statistical evaluation by means of logistic regression

In 33% of the patients examined, PR mRNA was detectable in the MNC fraction (Table 1 ). While PR mRNA expression does not correlate significantly with lymph node status, grading and tumor size, there is a distinct correlation with the menopausal status (Table 2; lymph node status (p = 0.95 for n = 368), grading (p = 0.29 for n = 368), tumor size (p = 0.88 for n = 368) and menopausal status (p = 0.007 for n = 368)). For those patients who had had no hormone therapy and had been recurrence-free for at least two years since diagnosis, there was no correlation of PR mRNA expression with the menopausal status either (Table 3; lymph node status (p = 0.99 for n = 44), grading (p = 0.41 for n = 44), tumor size (p = 0.67 for n = 44), and menopausal status (p = 0.33 for n = 44)). The evaluation shows furthermore that PR expression correlates with the absence of recurrences (Table 4; recurrence-free versus recurrence < 2 years: p = 0J7 for n = 280). In addition, chemotherapy and radiation correlate with the negative detection of PR mRNA expression; in contrast, hormone therapy, in particular tamoxifen administration, correlates with positive detection (Table 5; no hormones v. tamoxifen p = 0.003 for n = 340). When plotting the probability of recurrence, expressed as a percentage of those patients without recurrence, against the recurrence-free period for patients having positive PR mRNA detection ( ) and patients having negative PR mRNA detection ( — ) (cf. Kaplan- Meier diagram as in Figure 1 ), it is evident that patients having PR mRNA expression have significantly fewer recurrences than those patients who had no detectable PR mRNA expression (p = 0.05). Example 3: Bcl-2 mRNA

(a) Determination of the status of bcl-2 mRNA expression from disseminated cancer cells

5 ml of peripheral blood was taken in each case from 291 patients having surgically treated breast cancer. Mononuclear cells (MNC) were isolated by means of density gradient centrifugation. If present, disseminated cancer cells are in this cell fraction, too. Following examples 1 and 2 of WO 00/06702, fractions A (CD 45-positive isolates from MNC fraction), B (CD 45-positive isolates from the sieve flow-through) and C (sieve residue, cancer cell fraction having a disseminated cancer cell content of at least 1 :1) are obtained from the MNC fraction. RNA is then isolated from fractions A and C in a manner known per se and transcribed into cDNA by means of random hexamers and reverse transcriptase. Bcl-2-specific detection is then carried out by means of PCR and using the 5'-nuclease detection system known per se (TaqMan®). The following primer sequences and probe sequences were used: SEQ ID NO:4-6. For the evaluation, the ratio of bcl mRNA cell equivalents to GAPDH cell equivalents is formed for each of fractions A and C, and the ratio of the resulting quotients is in turn formed. If the ratio of fraction-C quotient to fraction-A quotient is greater than 1, bcl-2 mRNA is overexpressed. The cell equivalents are based on a cell standard. This cell standard is prepared by extracting RNA from a known number of cells (e.g. 2 x 10^"s) from a cell suspension of a carcinoma cell line expressing the particular parameter, and transcribing said RNA into cDNA. This cDNA is present in each quantitative analysis in the form of a dilution series (e.g. 6 dilution levels) and serves as a reference system.

As a control, blood from healthy individuals was analyzed accordingly.

(b) Cancer-related information about the clinical status of the individuals

Where present, information about the lymph node status, about grading, about tumor size, about the menopausal status, about the recurrence status, and about therapeutic measures previously carried out of the patients examined in (a) was compiled.

(c) Statistical evaluation by means of logistic regression In 33% of the patients examined, bcl-2 mRNA overexpression was detected in fraction C in comparison with fraction A (Table 1 ). There is no correlation of lymph node status, grading and tumor size with bcl-2 mRNA overexpression: there is no correlation of lymph node status (p=0.80 for n=153), grading (p=0.52 for n=153), tumor size (p=0.90 for n=153) and menopausal status (p=0.36 for n=153) with bcl-2 mRNA overexpression [sic]. This is in particular true for those patients who had no chemotherapy and were recurrence-free at the time of examination (lymph node status (p = 0.29 for n=58), grading (p=0.85 for n=58), tumor size (p=0.17 for n=58) and menopausal status (p=0.32 for n=58)). There is a correlation of overexpressed bcl-2 mRNA with the appearance of recurrences for chemotherapy and tamoxifen therapy, but not for radiotherapy (Table 7; recurrence v. no recurrence for radiation, no chemotherapy: p = 0J3 for n = 38, chemotherapy, no radiation: p = 0.06 for n = 27, tamoxifen: p = 0.01 for n = 66). In particular, detection of overexpressed bcl-2 mRNA correlates with the early appearance of recurrences (Table 4; recurrence-free versus recurrence < 2 years: p = 0.015 for n = 119). When plotting the probability of recurrence, expressed as the percentage of those patients without recurrence, against the recurrence-free period for patients having bcl-2 mRNA overexpression ( ) and patients having normal bcl-2 mRNA expression (— ) (cf. Kaplan-Meier diagram as in Figure 2), it is evident that patients having excessive bcl-2 mRNA expression have significantly more recurrences than those patients who had normal bcl-2 mRNA expression levels (p=0.013).

Example 4: ErbB2, CyclinB, CyclinD and bcl-2 mRNA

(a) Determination of the status of ErbB2, CylinB, CyclinD and bcl-2 mRNA expression from disseminated cancer cells

5 ml of peripherel blood was taken in each case from patients having surgically treated breast cancer. Mononuclear cells (MNC) were isolated by means of density gradient centrifugation. If present, disseminated cancer cells are in this cell fraction, too. Following examples 1 and 2 of WO 00/06702, fractions A (CD 45-positive isolates from MNC fraction), B (CD 45-positive isolates from the sieve flow-through) and C (sieve residue, cancer cell fraction having a disseminated cancer cell content of at least 1 :1) are obtained from the MNC fraction. RNA is then isolated from fractions A and C in a manner known per se and transcribed into cDNA by means of random hexamers and reverse transcriptase. bcl-2-specifιc detetcion was carried out as described in Example 3. ErbB2-, CyclinB- and CyclinD-specific detection is then carried out by means of PCR and using the 5^'-nuclease detection system known per se (TaqMan®). The following primer sequences and probe sequences were used: SEQ ID NO:10-18. For the evaluation, the ratio of ErbB2 (or CyclinB or CyclinD) mRNA cell equivalents to GAPDH cell equivalents is formed for each of fractions A and C, and the ratio of the resulting quotients is in turn formed. If the ratio of fraction- C quotient to fraction-A quotient is greater than 1 , ErbB2 mRNA (CyclinB/CyclinD) is overexpressed. The cell equivalents are based on a cell standard. This cell standard is prepared by extracting RNA from a known number of cells (e.g. 2 x 10^s) from a cell suspension of a carcinoma cell line expressing the particular parameter, and transcribing the said RNA into cDNA. This cDNA is present in each quantitative analysis in the form of a dilution series (e.g. 6 dilution levels) and serves as a reference system.

As a control, blood from healthy individuals was analyzed accordingly,

(b) Cancer related information about the clinical status of the individuals

(c) Statistical evaluation by means of logistic regression

In 22% of 306 patients analyzed, ErbB2 mRNA overexpression was detected in fraction C in comparison to fraction A. In 11% of 115 patients analyzed, CyclinB and CyclinD mRNA overexpression was detected in fraction C in comparison to fraction A. In patients with ErbB2 mRNA overexpression, additional overexpression of CyclinD mRNA was found in 29% of these patients whereas 18% showed low expression of CyclinD (Table 8; p=0.05). CyclinB mRNA overexpression was found in 30% of the patients with ErbB2 mRNA overexpression in contrast to 13% with low Cyclin B mRNA expression and ErbB2 mRNA overexpression (Table 8; p=0.001). Overexpression of bcl-2 was found in 43% of ErbB2 mRNA overexpressing patients in contrast to 25% with low bcl-2 mRNA expression and ErbB2 mRNA overexpression (Table 8; p=0.005).

There is no correlation of lymph node status, grading, tumor size or menopausal status with ErbB2-, CyclinB- and CyclinD-mRNA overexpression.

There is no correlation of overexpressed ErbB2 mRNA with the appearance of recurrence (Figure 3; recurrence v. no recurrence: p = 0.56). When plotting the probability of recurrence, expressed as the percentage of those patients without recurrence, against the recurrence-free period for patients (30/67) having ErbB2 mRNA overexpression ( ) and patients (119/239) having normal ErbB2 mRNA expression ( ) (cf. Kaplan-Meier diagram as in Figure 3), it is evident that patients having excessive ErbB2 mRNA expression have not significantly more recurrences than those patients who had normal ErbB2 mRNA expression levels.

There is also no correlation of combined overexpressed CyclinB and CyclinD mRNA with the appearance of recurrence (Figure 4; recurrence v. no recurrence: p = 0.97). When plotting the probability of recurrence, expressed as the percentage of those patients without recurrence, against the recurrence-free period for patients (6/13) having CyclinB and CyclinD mRNA overexpression ( ) and patients (44/102) having normal CyclinB and CyclinD mRNA expression ( ) (cf. Kaplan-Meier diagram as in Figure 4), it is evident that patients having excessive CyclinB and CyclinD mRNA expression have not significantly more recurrences than those patients who had normal CyclinB and CyclinD mRNA expression levels.

In contrast the combined overexpression of ErbB2, CyclinB and CyclinD mRNA correlates to the appearance of recurrence (Figure 5; recurrence v. no recurrence: p = 0.06). When plotting the probability of recurrence, expressed as the percentage of those patients without recurrence, against the recurrence-free period for patients (5/6) having combined overexpression of ErbB2, CyclinB and CyclinD mRNA ( ) and patients (29/68) having normal ErbB2 and CyclinB and CyclinD mRNA expression ( — — ) (cf. Kaplan-Meier diagram as in Figure 5), it is evident that patients having excessive ErbB2 and CyclinB and CyclinD mRNA expression have significantly more recurrences than those patients who had normal ErbB2 and CyclinB and CyclinD mRNA expression levels.

In particular, the combined overexpression of ErbB2, bcl-2, CyclinB and/or CyclinD mRNA correlates to the appearance of recurrence (Figure 6; recurrence v. no recurrence: p = 0.008). When plotting the probability of recurrence, expressed as the percentage of those patients without recurrence, against the recurrence-free period for patients (5/7) having combined overexpression of ErbB2, bcl-2, CyclinB and/or

CyclinD mRNA ( ) and patients (11/35) having normal expression of these parameters ( ) (cf. Kaplan-Meier diagram as in Figure 6), it is evident that patients with excessive mRNA expression have significantly more recurrences than those patients who had normal expression levels of these parameters.

In patients with combined ErbB2 and bcl-2 mRNA overexpression before therapy, Herceptin therapy of >1 year correlates with reduced detection of both ErbB2 and bcl- 2 mRNA.

Table 1 : Determination of the target status in the blood of n patients

Table 2: Comparison of target status and of parameters for the histological tumor classification according to the TNM system and of the menopausal status in all patients

Table 3: Comparison of target status and of parameters for the histological tumor classification according to the TNM system and of the menopausal status in patients recurrence-free for at least 2 years

Table 5: Comparison of target status and therapeutic treatments in patients recurrence-free for at least 2 years

Table 6: Comparison of target status and therapeutic treatments in patients recurrence-free for at least 2 years

Table 7: Comparison of bcl-2 mRNA overexpression and appearance of recurrences for various therapeutic treatments

Table 8: Overexpression of CyclinD, CyclinB and bcl-2 mRNA in patients with ErbB2 overexpression.

Claims

Patent Claims

1. Method for the clinical validation of a target from disseminated cancer cells, characterized in that for a population of individuals it is determined whether a status of the target determined in disseminated cancer cells of the individuals correlates with at least one cancer-related information about the clinical status of the individuals.

2. Method according to claim 1 , characterized in that the status of the target is determined in cell fractions derived from body fluids of the individuals, said cell fractions being enriched for disseminated cancer cells.

3. Method according to claim 2, characterized in that the cell fractions have a cancer cell to non-cancer cell proportion of more than 1:1.

4. Method according to any one of claims 1 to 3, characterized in that the prognostic relevance of the target is clinically validated.

5. Method according to any one of claims 1 to 3, characterized in that the therapeutic relevance of the target is clinically validated.

6. Method according to any one of claims 1 to 5, characterized in that the cancer-related information is the survival time and/or the cancer-free period of the individual.

7. Method according to any one of claim 1 to 5, characterized in that the cancer- related information is at least one therapeutic measure.

8. Method according to any of claims 1 to 7, characterized in that the individuals of the population have precancerous lesions.

9. Method according to any of claims 1 to 7, characterized in that the individuals of the population have a tumor.

10. Method according to any one of claims 1 to 7, characterized in that the individuals have a recurrence.

11. Method according to claim 9 or 10, characterized in that the individuals have not yet undergone systemic therapy.

12. Method according to claim 9 or 10, characterized in that the individuals have undergone systemic therapy.

13. Method according to claim 12, characterized in that the validation is carried out for a population of individuals who have undergone a first systemic therapy and at least for one further population of individuals who have undergone a further systemic therapy.

14. Method according to claim 13, characterized in that the first and each further systemic therapy is selected from radiotherapy, chemotherapy and adjuvant therapy.

15. Method according to any one of claims 8 to 14, characterized in that it is determined whether the target status determined in disseminated cancer cells of said individuals correlates with the survival time and/or the cancer-free period of the individuals.

16. Method for the clinical validation of a target from disseminated cancer cells, characterized in that it is determined whether a status of the target determined in disseminated cancer cells correlates with at least one further target status in said disseminated cancer cells.

17. Method according to claim 16, characterized in that said further target status correlates with with at least one item of cancer-related clinic information.

18. Method according to claim 17, characterized in that said further target status correlates with either (a) no tumor, (b) early tumor or (c) late tumor within an observation period in individuals with either (1) precancerous lesions, (2) having a manifest tumor prior to systemic therapy, or (3) patients after systemic therapy.

19. Method accoding to claim 18, characterized in that the obeservation period is at least 5 years.

20. Method according to any one of the preceding claims, characterized in that at least one splice variant of disseminated cancer cells is clinically validated as target.

21. Method for collecting a cancer-related dataset (profile) of an individual, characterized in that a) the individual is examined for disseminated cancer cells, b) optionally at least one aliquot of a sample derived from the body of the individual examined in a) is set aside in a substance library, c) at least one cancer-related information about the clinical status of the individual is provided, and d) the result obtained from the examination in a) and the information provided in c) and, optionally, the sample set aside in b) are linked to each other.

22. Method according to claim 21 , characterized in that steps c) and d) are repeated for the same individual at a later time.

23. Method according to claim 21 or 22, characterized in that steps a) and, optionally, b) and also d) are repeated for the same individual at a later time.

24. Method according to any one of claims 21 to 23, characterized in that the examination according to step a) includes determining the status of at least one target from the disseminated cancer cells.

25. Method according to any one of claims 21 to 24, characterized in that step b) is part of the method.

26. Method for building a database, in which the method according to any one of claims 21 to 25 is carried out for a multiplicity of individuals, and the datasets obtained are filed in a database.

27. Method for the functional validation of clinically validated targets from disseminated cancer cells, characterized in that at least one active substance is allowed to act ex vivo upon target-carrying disseminated cancer cells, and the biological reaction of said cancer cells is determined.

28. Method according to claim 27, characterized in that target-related active substances which have antisense function, antibody function or ribozyme function are used.

29. Method for treating cancer, in which

- the status of at least one validated target from disseminated cancer cells of an individual is determined and a therapy is chosen depending on said status.

30. Kit for target-related cancer therapy, which comprises

- agents for determining the status of at least one validated target in disseminated cancer cells of an individual;

- instructions for the therapy taking account of the determined status.

31. Use of at least one agent suitable for chemotherapie and of at least one agent suitable for radiotherapy in the manufacture of a medicament for treating individuals having disseminated cancer cells expressing CK 20 mRNA.

32. Use of at least one agent suitable for hormone therapie in the manufacture of a medicament for treating individuals having disseminated cancer cells expressing progesterone receptor mRNA.

33. Use of at least one agent suitable for radiotherapie in the manufacture of a medicament for treating individuals having disseminated cancer cells overexpressing bcl-2 mRNA.

34. Use of Herceptin in the manufacture of a medicament for treating individuals having disseminated cancer cells overexpressing ErbB2 mRNA.

35. Use of Herceptin in the manufacture of a medicament for treating individuals having disseminated cancer cells overexpressing ErbB2 mRNA and at least one further mRNA selected from bcl-2, CyclinB and CyclinD.