Detection of carcinoma metastases by nucleic acid amplification

Description Submit all comments and votes

BACKGROUND OF THE INVENTION

1. Field of the Invention

The present invention relates to the fields of molecular biology and oncology and provides novel methods for the detection of carcinoma metastases by nucleic acid amplification. In a preferred embodiment, carcinoma metastases are identified in hematopoietic tissues by detection of normal non-hematopoietic RNA expressed by the metastatic carcinoma cells. Detection of non-hematopoietic RNA sequences indicates the presence of metastatic disease. The methods have applications in the diagnosis, staging, and monitoring of carcinoma patients.

2. Description of Related Art

Recent advances in cancer therapeutics have demonstrated the need for more sensitive staging and monitoring procedures to ensure initiation of appropriate treatment, to define the end points of therapy and to develop and evaluate novel treatment modalities and strategies. In the management of carcinoma patients, the choice of appropriate initial treatment depends on accurate assessment of the stage of the disease. Patients with limited or regional disease generally have a better prognosis and are treated differently than patients who have distant metastases (Minna et al., 1989, Cancer Principals and Practices of Oncology, DeVita et al. ed. Lippincott, Philadelphia pp. 591-705, which is incorporated herein by reference). However, conventional techniques to detect these metastases are not very sensitive.

For example, the prognosis and therapeutic management of both major histological subgroups of lung cancer (small cell and non-small cell) depend upon the stage of disease activity at the time of diagnosis (Green, 1989, Lung Cancer 5:178-185, which is incorporated herein by reference). Patients with non-small lung cancers cell (NSCLC) comprise approximately 75% of lung cancers. This histological subgroup of lung cancer has been considered relatively resistant to chemotherapy. However, NSCLC are often curable by surgical resection (and occasionally by radiation therapy) in patients with stages I,II, or IIIA disease who do not have occult distant metastases.

Unfortunately, conventional staging procedures to detect metastatic disease are not very sensitive. Approximately 25% to 30% of stage I NSCLC patients are not cured by primary tumor resection because they have metastases that are not identified by standard methods during preoperative staging. The development of more sensitive techniques to detect metastases could identify those NSCLC patients who will not be cured by local surgical rumor resection, who would benefit from the administration of effective systemic therapies. Similarly, more sensitive methods to detect metastases in other types of carcinomas would identify patients who will not be cured by local therapeutic measures, for whom effective systemic therapies would be more appropriate.

The inadequacy of current staging methods also adversely effects the management of the other major histological type of lung cancer, small cell lung carcinoma (SCLC). In contrast to NSCLC, SCLC is very sensitive to chemotherapy and radiation therapy but is generally believed to be incurable by surgery alone since these minors have usually metastasized to distant sites at the time of diagnosis. Approximately 25% to 35% of limited disease patients who achieve complete remissions with therapy have durable remissions, and two years event-free survival, following treatment. Current staging procedures cannot distinguish those who will have earlier relapses despite achieving initial complete remission. Most SCLC patients who achieve complete remissions have minimal residual disease (MRD) which cannot be detected by conventional methods. More sensitive methods to detect metastases are needed for identifying limited disease patients at high risk for early rumor recurrence who may benefit from additional systemic therapy.

Immunocytological procedures have been used to detect cancer cells in peripheral blood and bone marrow specimens unsuspected on the basis of conventional morphological evaluations (Sobol et al., 1982, Clin. Immunopathol. 24:139-144, and Sobol et al., 1985, Cancer 56:2005-2010, incorporated herein by reference). Immunohistochemical and immunofluorescence techniques have been used to identify antigens expressed by carcinomas that are not expressed by hematopoietic tissues (Sobol et al., 1986, Cancer Research 46:4746-4750, and Sobol et al., 1987, Ann. Intern. Med. 105:698-700). Several investigators have employed monoclonal antibody immunocytology to detect bone marrow metastases in carcinoma patients not identified by standard morphological examinations (Cannon et al., 1988, Eur. J. Cancer Clin. Oncol. 24:147-150; Berendsen et al., 1988, J. Clin. Pathol. 41:273-276; and Stahel et al., 1985, J. Clin. Oncol 3:455-461). However, immunocytological and standard morphological evaluations can reliably detect only 1% to 5% malignant cells in a mixed population with normal hematopoietic cells (Wright et al., 1987, J. Clin. Oncology 5:735-741).

Nucleotide amplification techniques provide rapid and sensitive methods for detecting specific nucleotide sequences (Mullis et al., 1986, Cold Spring Harbor Symposium Quant. Biol. 51:263-273, and Saiki et al., 1988, Science 239:487-491, which are incorporated herein by reference). Cell mixing experiments have demonstrated that polymerase chain reaction (PCR) analysis can identify as few as 1:10.sup.4 or 1:10.sup.5 cells that contain a target gene sequence (Kawasaki et al., 1987, Proc. Natl. Acad. Sci. USA 85:5698-5702, and Crescenzi et al., 1988, Proc. Natl. Acad. Sci. USA 85:4869-4873, which are incorporated herein by reference).

PCR has been employed to detect minimal residual disease activity in patients with hematopoietic malignancies (Fey et al., 1991, Eur. J. Cancer 27:89-94, and PCT Patent Publication No. WO 89/087 17). These methods rely on the identification of abnormal nucleotide sequences resulting from recurring chromosome translocations which characterize the hematological malignancy. Primers flanking the chromosome break points are employed to amplify the aberrant nucleotide sequences which result from the translocation event. In contrast, recurring chromosome translocations are not a common feature of carcinomas. Some carcinomas are characterized by aberrant oncogene or tumor suppressor gene nucleotide sequences (Cooper, 1990, Oncogenes Jones and Burnlett Publishers). However, these abnormal nucleotide sequences are either too diverse, poorly characterized, or infrequent to serve as targets for a practical, generally applied nucleic acid amplification procedure to detect metastatic carcinomas. Novel methods are needed to exploit the sensitivity of nucleic acid amplification procedures to detect metastatic carcinoma disease activity. The present invention meets these needs.

SUMMARY OF THE INVENTION

The invention provides a method for detecting carcinoma metastases in body tissues and fluids that comprise the steps of: (a) treating a sample containing nucleic acid from the cells of the body tissues and fluids under conditions for amplifying a target carcinoma associated sequence in an amplification reaction mixture that comprises a primer pair for specifically amplifying the target carcinoma associated sequence, to provide an amplified sequence if the target carcinoma associated sequence is present, wherein the target carcinoma associated sequence is indicative of carcinoma metastases in the body tissues and fluids; and (b) determining if amplification has occurred.

In another aspect, the invention provides oligonucleotide primers and probes for amplifying and detecting metastatic disease in body tissues and fluids, wherein the primers are suitable for amplifying a target carcinoma associated sequence, which sequence is preferentially expressed in carcinoma tumor cells and not in the body tissues and fluids to be analyzed for detecting carcinoma metastases.

In another aspect, the invention provides kits for detecting metastatic disease in body tissues and fluids, the kits comprising: (a) a primer pair for amplifying a target carcinoma associated sequence, which sequence is preferentially expressed in carcinoma cells and not in resident cells normally present in the body tissues and fluids.

In another aspect, the invention provides a method for identifying a carcinoma associated RNA sequence suitable as a cancer marker for detecting carcinoma metastases in body tissues or fluids comprising the steps of: (a) isolating negative control mRNA from non-carcinoma cells; (b) isolating positive control mRNA from carcinoma cells; (c) reverse transcribing and amplifying the negative and positive control mRNAs in separate amplification reaction mixtures comprising a primer pair for specifically amplifying a candidate carcinoma associated sequence, wherein the sequence is normally expressed by epithelial cells but not by the non-carcinoma cells; and (d) determining if amplification has occurred in the positive control sample and if amplification has failed in the negative control sample.

DETAILED DESCRIPTION

The present invention provides a method for detecting carcinoma associated nucleic acids in body tissues and fluids. In the preferred embodiment of the invention, the nucleic acids to be detected are carcinoma associated RNAs.

The invention is useful for detecting and monitoring carcinoma patients. In another aspect, the detection of carcinoma metastases has applicability in assessing the suitability of remission bone marrow specimens for carcinoma therapies incorporating autologous bone marrow transplantation. The availability of the present methods for detecting metastatic disease has utility in redefining the staging criteria for carcinomas and aide in determining the most appropriate type of initial therapy. The improved methods to detect metastases permit more precise documentation of complete remissions and early relapses. This information provides useful guidance for making decisions as to whether therapy should be continued, reinstituted or ceased and potentially results in more appropriate overall therapy for carcinoma patients. These novel methods provide means towards more effective management of carcinomas.

According to the present methods, RNA is extracted from cells in body tissues or fluids, for example, hematopoietic tissue such as bone marrow or peripheral blood and incubated with reverse transcriptase and deoxyribonucleoside triphosphates to generate cDNA. Subsequently, amplification procedures are employed to detect gene sequences expressed by carcinoma cells but not by the resident non-carcinoma cells normally present in the sample body tissue or fluid. In a preferred embodiment, PCR analysis of carcinoma associated mRNA is used to detect carcinoma cells in bone marrow specimens. Following reverse transcription, DNA polymerase and the up-stream and down-stream primers for the target sequences of interest are added to the reaction mixture to amplify the target gene sequence. In one embodiment, following 20-40 cycles of amplification using PCR methods, the reaction mixture is extracted with chloroform and the aqueous phase is electrophoresed in an agarose gel. The gel is stained with ethidium bromide and photographed to determine the presence of the target sequence. Oligonucleotide probes may be used to unequivocally identify the target sequence using Southern or dot blot methodologies.

In contrast to prior methods for cancer detection, the target nucleic acid is not necessarily an oncogene mRNA product. Carcinomas are non-hematopoietic cancers and no common recurrent translocations or uniform gene aberrations have been identified for reproducibly identifying metastasized carcinomas by nucleic acid amplification methods. Conventional methods are useful for detecting metastases where 1-5% of the cells analyzed display the cytological characteristics of cancer cells.

The present invention for detecting carcinoma metastases also provides methods for identifying carcinoma associated nucleic acid sequences useful as cancer markers. As used herein, the term "carcinoma" refers to malignancies of epithelial or endocrine tissues including respiratory system carcinomas, gastrointestinal system carcinomas, genitourinary system carcinomas, testicular carcinomas, breast carcinomas, prostatic carcinomas, endocrine system carcinomas, and melanomas.

As used herein, the term "carcinoma associated sequences" or "carcinoma associated RNA sequences" refers to nucleic acid sequences expressed by carcinoma cells, that are not expressed by the resident cells normally present in the sample body tissues and fluids. As used herein, generally, carcinoma associated sequences are RNAs.

The present invention provides methods for identifying particular target nucleic acids and amplifying those nucleic acids, for detecting carcinoma metastases in body tissues or fluids such as hematopoietic tissues (bone marrow, peripheral blood, and lymph nodes), and pleural effusions. The target nucleic acids, as described herein, are carcinoma associated RNA transcripts produced in cancer cells, as well as in healthy cells from which the tumors arise. However, according to the invention, the presence of metastatic disease is determined by amplifying and detecting these target nucleic acids in cells of tissues or fluids which do not normally express the target genes. In the preferred embodiment of the invention, the target genes are not normally expressed in hematopoietic cells.

To determine whether or not a particular carcinoma has metastasized, the specific type of carcinoma can be considered for determining suitable target nucleic acids to be detected. For example, genes expressed by carcinomas, that are suitable for detecting metastases in bone marrow and peripheral blood specimens, would include genes encoding epithelial antigens or neuroendocrine antigens that are not expressed by hematopoietic cells. Some target nucleic acids, particularly those encoding epithelial antigens, such as the antigen recognized by the monoclonal antibody KS 1/4 (Bumol et al., 1988, Hybridoma 7(4): 407-415), are useful for detecting a broad spectrum of metastasized carcinomas.

Criteria for selecting target genes for analysis include gene expression by a large percentage of carcinoma cells and the absence of expression by hematopoietic elements. Target genes include but are not limited to chromogranin A (chromo A ), neuron specific enolase, calcitonin, bombesin, neural cell adhesion molecules (NCAM), synaptophysin (synapto), L-dopa decarboxylase, neurophysin I and II (neuro I, neuro II), parathyroid related hormone of malignancy, selected SCLC antigens defined by monoclonal antibodies, and the pan-carcinoma antigens recognized by the monoclonal antibody KS1/4. It will be obvious to one of ordinary skill in the art that the suitability of any particular target gene for use in the present methods depends on the particular primers, samples, and conditions employed. Methods for assessing the suitability of a target gene as a cancer marker are disclosed herein and demonstrated in the examples.

The sensitivity of the present methods distinguish the invention from prior methods for detecting metastasized tumor cells. Previous methods include histological and serum chemistry analyses, physical examinations, bone scans, and X-rays. The examples disclosed herein demonstrate PCR amplification for detecting target gene products. However, any of a number of amplification methods are equally suitable for practicing the invention.

The term "amplification reaction system" refers to any in vitro means for multiplying the copies of a target sequence of nucleic acid. Such methods include but are not limited to polymerase (PCR), DNA ligase, (LCR), Q.beta. RNA replicase, and RNA transcription-based (TAS and 3SR) amplification systems.

The term "amplifying" which typically refers to an "exponential" increase in target nucleic acid is being used herein to describe both linear and exponential increases in the numbers of a select target sequence of nucleic acid.

The term "amplification reaction mixture" refers to an aqueous solution comprising the various reagents used to amplify a target nucleic acid. These include enzymes, aqueous buffers, salts, amplification primers, target nucleic acid, and nucleoside triphosphates. Depending upon the context, the mixture can be either a complete or incomplete amplification reaction mixture.

The systems described below are practiced routinely by those of skill in the relevant art. They have been described in detail by others and are summarized below. This invention is not limited to any particular amplification system. As other systems are developed, those systems may benefit by practice of this invention. A recent survey of amplification systems was published in Bio/Technology 8:290-293, April 1990, incorporated herein by reference. The following four systems are described below for the convenience of those not familiar with amplification systems and to provide an understanding of the breadth of the present invention.

Amplification of DNA by PCR is disclosed in U.S. Pat. Nos. 4,683,195 and 4,683,202 (both of which are incorporated herein by reference). Methods for amplifying and detecting nucleic acids by PCR using a thermostable enzyme are disclosed in US. Pat. No. 4,965,188, which is incorporated herein by reference.

PCR amplification of DNA involves repeated cycles of heat-denaturing the DNA, annealing two oligonucleotide primers to sequences that flank the DNA segment to be amplified, and extending the annealed primers with DNA polymerase. The primers hybridize to opposite strands of the target sequence and are oriented so that DNA synthesis by the polymerase proceeds across the region between the primers, effectively doubling the amount of the DNA segment. Moreover, because the extension products are also complementary to and capable of binding primers, each successive cycle essentially doubles the amount of DNA synthesized in the previous cycle. This results in the exponential accumulation of the specific target fragment, at a rate of approximately 2 per cycle resulting in an accumulation of 2.sup.n, where n is the number of cycles.

In the disclosed embodiment, Taq DNA polymerase is preferred although this is not an essential aspect of the invention. Taq polymerase, a thermostable polymerase, is active at high temperatures. Methods for the preparation of Taq are disclosed in U.S. Pat. No. 4,889,818 and incorporated herein by reference. Taq polymerase is available from Perkin Elmer Cetus Instruments (PECI). However, other thermostable DNA polymerases isolated from other Thermus species or non Thermus species (e.g., Thermus thermophilus or Thermotoga maritima), as well as non-thermostable DNA polymerase such as T4 DNA polymerase, T7 DNA polymerase, E. coli DNA polymerase I, or the Klenow fragment of E. coli, can also be used in PCR. Methods for providing thermostable DNA polymerases are provided in Ser. No. 08/148,133, which is a continuation of Ser. No. 4,55,967, filed Dec. 22, 1989, now abandoned; U.S. Pat. No. 5,374,553; U.S. Pat. No. 5,405,774; and U.S. Pat. No. 5,455,170; and Ser. No. 550,490, filed Sep. 28, 1990, now abandoned, which are all incorporated herein by reference.

The nucleoside-5'-triphosphates utilized in the extension process, typically dATP, dCTP, dGTP, and dTTP, are present in total concentration typically ranging from 0.05 mM to 0.5 mM during the extension reaction, although preferably the concentration is between 0.1 mM and 0.2 mM.

As used herein, the term "primer" refers to an oligonucleotide capable of acting as a point of initiation of DNA synthesis when annealed to a nucleic acid template under conditions in which synthesis of a primer extension product is initiated, i.e., in the presence of four different nucleotide triphosphates and a DNA polymerase in an appropriate buffer ("buffer" includes. pH, ionic strength, cofactors, etc.) and at a suitable temperature.

The choice of primers for use in PCR determines the specificity of the amplification reaction. Primers used in the present invention are oligonucleotides, usually deoxyribonucleotides several nucleotides in length, that can be extended in a template-specific manner by the polymerase chain reaction. The primer is sufficiently long to prime the synthesis of extension products in the presence of the agent for polymerization and typically contains 10-30 nucleotides, although that exact number is not critical to the successful application of the method. Short primer molecules generally require cooler temperatures to form sufficiently stable hybrid complexes with the template.

Synthetic oligonucleotides can be prepared using the triester method of Matteucci et al., 1981, J. Am. Chem. Soc. 103:3185-3191. Alternatively automated synthesis may be preferred, for example, on a Biosearch 8700 DNA Synthesizer using cyanoethyl phosphoramidite chemistry.

For primer extension to occur, this primer must anneal to the nucleic acid template. Not every nucleotide of the primer must anneal to the template for extension to occur. The primer sequence need not reflect the exact sequence of template. For example, a non-complementary nucleotide fragment may be attached to the 5' end of the primer with the remainder of the primer sequence being complementary to the template. Alteratively, non-complementary bases can be interspersed into the primer, provided that the primer sequence has sufficient complementarily with the template for annealing to occur and allow synthesis of a complementary DNA strand.

Due to the enormous amplification possible with the PCR process, small levels of DNA carryover from samples with high DNA levels, positive control templates or from previous amplifications can result in PCR product, even in the absence of added template DNA. If possible, all reaction mixes are set up in an area separate from PCR product analysis and sample preparation. The use of dedicated or disposable vessels, solutions, and pipettes (preferably positive displacement pipettes) for RNA/DNA preparation, reaction mixing, and sample analysis will minimize cross contamination. See also Higuchi and Kwok, 1989, Nature, 339:237-238 and Kwok, and Orrego, in: Innis et al. eds., 1990 PCR Protocols: A Guide to Methods and Applications, Academic Press, Inc., San Diego, Calif., which are incorporated herein by reference.

One particular method for minimizing the effects of cross contamination of nucleic acid amplification is described in U.S. Ser. No. 609,157, filed Nov. 2, 1990, now abandoned, which is incorporated herein by reference. The method involves the introduction of unconventional nucleotide bases, such as dUTP, into the amplified product and exposing carryover product to enzymatic and/or physical-chemical treatment to render the product DNA incapable of serving as a template for subsequent amplifications. For example, uracil-DNA glycosylase will remove uracil residues from PCR product containing that base. The enzyme treatment results in degradation of the contaminating carryover PCR product and serves to "sterilize" the amplification reaction.

Amplification systems such as PCR require a target nucleic acid in a buffer compatible with the enzymes used to amplify the target The target nucleic acid can be isolated from a variety of biological materials including tissues, body fluids, feces, sputum, saliva, and the like. In the preferred embodiment of the invention, the target nucleic acid is a carcinoma associated RNA sequence and the sample to be tested for the presence of the target nucleic acid is contained in a hematopoietic tissue sample, for example, bone marrow aspirate biopsies, peripheral blood, lymph node cell biopsies or aspirates. Other samples suitable for analysis include but are not limited to pleural fluid, ascites, and cerebrospinal fluid.

To amplify a target nucleic acid sequence in a sample, the sequence must be accessible to the components of the amplification system. In general, this accessibility is ensured by isolating the nucleic acids from a crude biological sample. A variety of techniques for extracting nucleic acids from biological samples are known in the art. For example, see those described in Sambrook et at., Molecular Cloning: A Laboratory Manual (New York, Cold Spring Harbor Laboratory, 1989); Arrand, Preparation of Nucleic Acid Probes, in pp. 18-30, Nucleic Acid Hybridization: A Practical Approach (Ed Hames and Higgins, IRL Press, 1985); or, in PCR Protocols, Chapters 18-20 (Innis et al., ed., Academic Press, 1990), which are all incorporated herein by reference.

In general, the nucleic acid in the sample will be a sequence of RNA or DNA. RNA is prepared by any number of methods; the choice may depend on the source of the sample and availability. Methods for preparing RNA are described in Davis et al., 1986, Basic Methods in Molecular Biology, Elsevier, N.Y., Chapter 11; Ausubel et al., 1987, Current Protocols in Molecular Biology, Chapter 4, John Wiley and Sons, N.Y.; Kawasaki and Wang, 1989, PCR Technology, ed. Erlich, Stockton Press N.Y.; Kawasaki, 1990, PCR Protocols: A Guide to Methods and Applications, Innis et al. eds. Academic Press, San Diego; and Wang and Mark, 1990, PCR Protocols: A Guide to Methods and Applications, Innis et al. eds. Academic Press, San Diego; all of which are incorporated herein by reference. Chomczynski and Sacchi, 1987, Anal. Biochem 162:156-159, which is incorporated herein by reference, provides a single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Those of skill in the an will recognize that whatever the nature of the nucleic acid, the nucleic acid can be amplified merely by making appropriate and well recognized modifications to the method being used.

It is preferred, but not essential that the thermostable DNA polymerase is added to the reaction mix after both the primer and the template are added. Alternatively, for example, the enzyme and primer are added last, or the MgCl.sub.2, or template plus MgCl.sub.2 are added last. It is generally desirable that at least one component, that is essential for polymerization, not be present, until such time as the primer and template are both present and the enzyme can bind to and extend the desired primer/template substrate. This modification of PCR is referred to as "hot start" and is described in U.S. Pat. No. 5,411,876, which is incorporated herein by reference.

Those skilled in the an will know that the PCR process is most usually carded out as an automated process with a thermostable enzyme. In this process, the reaction mixture is cycled through a denaturing temperature range, a primer annealing temperature range, and an extension temperature range. Generally, the annealing and extension temperature ranges overlap, and consequently, PCR is often practiced as a two-step cycling reaction comprising a denaturing step and an annealing/extension step. A machine specifically adapted for use with a thermostable enzyme is disclosed more completely in EP No. 236,069, which is incorporated herein by reference, and is commercially available from PECI.

The ligase chain reaction is described in PCT Patent Publication No. WO 89/09835, which is incorporated herein by reference. The process involves the use of ligase to join oligonucleotide segments that anneal to the target nucleic acid. Ligase chain reaction (LCR) results in amplification of an original target molecule and can provide millions of copies of product DNA. Consequently, the LCR results in a net increase in double-stranded DNA. The present detection methods are applicable to LCR, as well as PCR. LCR requires an oligonucleotide probe for detecting the product DNA.

Another amplification scheme exploits the use of the replicase from the RNA bacteriophage Q.beta.. In this amplification scheme, a modified recombinant bacteriophage genome with a sequence specific for the targeted sequence is initially hybridized with the nucleic acid to be tested. Following enrichment of the duplexes formed between the bacteriophage probe and the nucleic acid in a sample, Q.beta. replicase is added, which, upon recognizing the retained recombinant genome, begins making large numbers of copies.

The Q.beta. system does not require primer sequences and there is no heat denaturation step as with the PCR and LCR amplification systems. The reaction occurs at one temperature, typically 37.degree. C. The preferred template is a substrate for the Q.beta. replicase, midvariant-1 RNA. A very large increase in the templates is achieved through the use of this system. A review of this amplification system can be found in the International Patent Application Pub. No. WO 87/06270 and in Lizardi et at., 1988, Bio/Technology 6:1197-1202.

The 3SR system is a variation of an vitro transcription based amplification system. A transcription-based amplification system (TAS) involves the use of primers that encode a promoter to generate DNA copies of a target strand and the production of RNA copies from the DNA copies with an RNA polymerase. See, e.g., Example 9B of U.S. Pat. No. 4,683,202 and EP No. 310,229. The 3SR System is a system which uses three enzymes to carry out an isothermal replication of target nucleic acids.

The system begins with a target of single-stranded RNA to which a T7 RNA DNA primer is bound. By extension of the primer with reverse transcriptase, a cDNA is formed, and RNAseH treatment frees the cDNA from the heteroduplex. A second primer is bound to the cDNA and a double stranded cDNA is formed by DNA polymerase (i.e., reverse transcriptase) treatment. One (or both) of the primers encodes a promoter, i.e., the promoter for T7 RNA polymerase, so that the double-stranded cDNA is transcription template for T7 RNA polymerase.

Transcription competent cDNAs yield antisense RNA copies of the original target. The transcripts are then convened by the reverse transcriptase to double standard cDNA containing double-stranded promoters, optionally on both ends in an inverted repeat orientation. These DNAs can yield RNAs, which can reenter the cycle. A more complete description of the 3SR system can be found in Guatelli et al., 1990, Proc. Natl. Acad. Sci. USA 87:1874-1878, and EP No. 329,822, both of which are incorporated herein by reference. The TAS system is also described in Gingeras et in Innis et al. eds., 1990, PCR Protocols, Academic Press, San Diego, which is incorporated herein by reference.

In the process described herein, a sample is provided which contains, or is suspected of containing, a particular oligonucleotide sequence of interest, the "target nucleic acid." The target may be RNA or DNA or an RNA/DNA hybrid. The target may be single stranded or double stranded. According to the present invention, the target nucleic acid is a carcinoma associated sequence. For example, in one aspect of the invention, a carcinoma may be attributable to an integrated virus, such as human papilloma virus (HPV), and cervical cancer cells would contain HPV DNA sequences. Detection of HPV DNA sequences in non-cervical cells, such as bone marrow, by the present methods provides, evidence of metastatic disease. Target preparation will be carried out in a manner appropriate for the particular amplification process to be implemented. For example, in a PCR method where the target nucleic acid is single-stranded, such as mRNA, the target may be first reverse-transcribed into cDNA, prior to amplification.

Although the PCR procedure described above assumed a double-stranded target, this is not a necessity. After the fast amplification cycle of a single-stranded DNA target, the reaction mixture contains a double-stranded DNA molecule consisting of the single-stranded target and a newly synthesized complementary strand. Similarly, following the fast amplification cycle of an RNA/cDNA target, the reaction mixture contains a double-stranded cDNA molecule and a duplicate of the original RNA/cDNA target. At this point, successive cycles of amplification proceed as described above. In the present methods, the target of amplification is a single-stranded RNA, and the fast amplification cycle is the reverse transcription step.

Methods for reverse transcribing RNA into cDNA are well known and described in Sambrook et al., supra. Alteratively, preferred methods for reverse transcription utilize thermostable DNA polymerases. These methods are described in U.S. Pat. No. 5,322,770, and WO 90/07641 filed Dec. 21, 1990, incorporated herein by reference U.S. Pat. No. 5,322,770 describes a procedure for coupled reverse transcription/amplification of an RNA template using a thermostable DNA polymerase.

Target genes are selected from those that are preferentially expressed in epithelial tissues and not in hematopoietic tissues. Amplification primers are preferably designed to hybridize to exons whose sequences are adjacent in mRNA. In this manner, background amplification of genomic sequences is minimized and readily distinguished from target amplification by size.

According to the invention, for selecting target genes suitable as cancer markers, primers are tested using available characterized carcinoma cell lines as positive controls, for example, from the American Type Culture Collection, Rockville, Md., and representative body tissues and fluid specimens, for example, bone marrow aspirates, from individuals without carcinomas, as negative controls in amplification reactions. For example, the existence of pseudogenes is determined so that a false positive result will not be obtained if genomic DNA contaminates a target sample. When a genomic pseudogene is present methods for avoiding pseudogene amplification will be obvious to one of ordinary skill in the an by, for example, selecting a different target sequence region to be amplified or altering the specificity of the reaction by varying primer length or cycling parameters.

Although the target carcinoma associated RNA sequences are preferentially expressed in epithelial tissues, the body fluids or tissues to be evaluated may express target sequences at a low basal level. Because of the sensitivity of the PCR method, even basal level expression can compromise a test (Chelly et at., 1988, Nature 330:858-860). Consequently, the identification of carcinoma associated sequences for detecting carcinoma metastases in the selected body tissues and fluids preferably includes screening amplification and detection evaluations to ascertain (1) the absence of (failure to amplify and detect) the target carcinoma associated sequence and non-malignant cells from representative samples of body tissues and fluids to be tested for carcinoma metastases and (2) the presence of (ability to amplify and detect) the target carcinoma associated sequences and representative carcinoma cells. Accordingly, it is preferred that normal bone marrow specimens and specimens from patients with hematopoietic cancers, as well as normal peripheral blood samples are analyzed to determine preferred primers and targets for detecting metastasized carcinomas. If necessary, one of skill in the art can readily modify the amplification and/or detection methods to distinguish between a low basal level of expression in a non-carcinoma cell, and a positive amplification from a carcinoma cell sample. For example, the amount of sample, cycling parameters, and detection scheme are modified as needed to reduce the likelihood of a false positive.

Body tissues and fluids suitable for detecting carcinoma metastases include, but are not limited to, those which are currently evaluated to detect metastatic dissects by standard, less sensitive cytological methods: bone marrow aspirates and biopsies, pleural effusions, ascites, cerebrospinal fluid, lymph node aspirates and biopsies, and peripheral blood.

Candidate carcinoma associated sequences include but are not limited to RNA sequences which encode proteins previously shown by conventional protein detection assays to be preferentially expressed by carcinoma cells and not by cells in the body tissues and fluids that will be evaluated for carcinoma metastases. Target sequences for the detection of metastasized carcinomas are briefly described below. For any particular target sequence, it will be obvious to one of ordinary skill in the art to select primers for amplification in accordance with the description of the methods provided herein.

KS 1/4 pan-carcinoma antigen is a monoclonal antibody defined antigen expressed by most carcinomas, including small cell and non-small cell lung cancers (Perez and Walker, 1990, J. Immunol. 142:3662-3667, and Bumal, 1988, Hybridoma 7(4):407-415, which are incorporated herein by reference). The frequency of expression in SCLC or KS 1/4 is approximately 80%. This gene is also highly expressed in other carcinomas.

Synaptophysin and bombesin/gastrin releasing peptide (GRP) are both neuroendocrine peptides and both are frequently expressed by small cell carcinomas (approximately 60%-80%) (Edbrooke et al., EMBO J. 4:715-724; Kayser et al., 1988, Pathology Research and Practice 183(4):412-417; and Spindelet, al., 1986, Proc. Natl. Acad. Sci. USA 83:19-23, which are incorporated herein by reference). L-dopa decarboxylase is also expressed by a high percentage of small cell lung carcinoma cells (approximately 80%) (Gazdar et al., 1988, Cancer Research 48:4078-4082, which is incorporated herein by reference).

Neuron specific enolase is expressed by a high percentage of small lung cell carcinomas (60-80%) (Kayser et at., 1988, Pathology Research Practice 143:412-417, which is incorporated herein by reference). Preferably, for practicing the present invention, primers for this gene should avoid sequence homologies with other forms of enolases that are frequently expressed by non-neuroendocrine tissues.

Parathyroid related hormone (PRH) may be expressed by more than 50% of small lung cell carcinomas, as well as in several other carcinomas. PRH is also expressed by hematopoietic malignancies associated with HTLV I infection. In a preferred embodiment of the invention, primers that correspond to amino acids 35 to 139 are selected to avoid regions of homology with parathyroid hormone and include all known forms of PRH (Suva, 1989, Gene 77(1):95-105; and Martin et al., 1989, Recent Progress in Hormone Research 45:467-506).

Calcitonin may be more variably expressed by small cell lung carcinomas (Jonas et al., 1985, Proc. Natl. Acad. Sci. USA 82:1994-1998, and Russell et al., 1990, Mol. Cell Endocrin. 71(1):1-12, which are incorporated herein by reference). Calcitonin and/or the calcitonin gene related peptide (CGRP) are expressed by 40%-60% small cell lung carcinomas. In the disclosed example, primers are designed as consensus primers for amplifying both calcitonin and CGRP sequences to enhance the yield of cancers detected.

Chromogranin A is associated with secretory granules found in normal neuroendocrine cells and in neuroendocrine tumors (Konecki et al., 1987, JBC 262:17026-17030, and Sobol et al., 1986, Annals of Internal Medicine 105(5):698-700, which are incorporated herein by reference). The protein is expressed by 40%-60% small lung cell carcinomas. According to the preferred embodiment, primers are selected from the middle region of the molecule. The amino and carboxy terminal regions share homology with chromogranin B and C which may have less specificity for neuroendocrine cells.

Neurophysins are precursors of oxytocin and vasopressin which are expressed by approximately 30-50% of small cell carcinomas (Mohr et al., 1985, FEBS Letters 93:12-16; North et al., 1988, Cancer 62(7):1343-1347, and Kibbelaar et al., 1989, J. Pathology 159:23-28, which are incorporated herein by reference). Consensus primers to amplify both neurophysins I and II may be utilized to enhance the yield of small cell lung carcinomas detected.

Other candidate target sequences include but are not limited to ovarian carcinoma antigen (CA125) (Yu et al., 1991, Cancer Res. 51(2):468-475); prostatic acid phosphate (Tailor et at., 1990, Nuc. Acids Res 18( 16): 4928); prostate specific antigen (Henttu and Vihko, 1989, Biochemical and Biophys Res. Comm 160(2):903-910); melanoma-associated antigen p97 (Estin et al., 1989, J. Natl. Cancer. Instit. 81(6):445-446); melanoma antigen gp75 (Vijayasardahi et al., 1990, Z. Experimental Medicine 171(4):1375-1380 and high molecular weight melanoma antigen (Natali et 1987, Cancer 59:55-63). These publications are all incorporated herein by reference.

It will be obvious to one or ordinary skill in the art that positive control cell lines are necessary in an initial evaluation for amplifying target genes. For the study of any particular carcinoma, numerous cell lines are described in the literature and available thro