1999;59:3251-3256. Cancer Res   Jose M. Silva, Gemma Dominguez, Jose M. Garcia, et al.   Clinicopathological CorrelationsPresence of Tumor DNA in Plasma of Breast Cancer Patients:

  Updated version

  http://cancerres.aacrjournals.org/content/59/13/3251

Access the most recent version of this article at:

   

   

  Cited Articles

  http://cancerres.aacrjournals.org/content/59/13/3251.full.html#ref-list-1

This article cites by 51 articles, 25 of which you can access for free at:

  Citing articles

  http://cancerres.aacrjournals.org/content/59/13/3251.full.html#related-urls

This article has been cited by 48 HighWire-hosted articles. Access the articles at:

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  .permissions@aacr.orgDepartment at

To request permission to re-use all or part of this article, contact the AACR Publications

Research. on February 5, 2014. © 1999 American Association for Cancercancerres.aacrjournals.org Downloaded from

Research. on February 5, 2014. © 1999 American Association for Cancercancerres.aacrjournals.org Downloaded from

[CANCER RESEARCH 59, 3251–3256, July 1, 1999]

Presence of Tumor DNA in Plasma of Breast Cancer Patients:Clinicopathological Correlations1

Jose M. Silva, Gemma Dominguez, Jose M. Garcia, Rocio Gonzalez, Maria J. Villanueva, Fatima Navarro,Mariano Provencio, Santiago San Martin, Pilar Espana, and Felix Bonilla2

Department of Medical Oncology, Clinica Puerta de Hierro [J. M. S., G. D., J. M. G., R. G., M. J. V., F. N., M. P., P. E., F. B.], and Department of Gynecology, Hospital SantaCristina [S. S. M.], 28035 Madrid, Spain

ABSTRACT

Using different molecular techniques, DNA has been detected in theplasma of cancer patients with various types of tumors. We undertook thepresent study to investigate the presence of plasma DNA, before mastec-tomy, in patients with breast cancer at diagnosis and to analyze theclinicopathological spectrum of this subgroup of patients with respect topatients without DNA with tumor characteristics. We studied 62 patientswith breast cancer, who were selected sequentially after mastectomy anddiagnosis of breast carcinomas. Genomic DNA extracted from tumor andnormal tissues, normal blood cells, and plasma was used for molecularstudies. Alterations in polymorphic markers selected because they hadbeen found to show a high rate of alterations in breast cancer in previousstudies (D17S855, D17S654, D16S421, TH2, D10S197, and D9S161), aswell as mutations in the p53 gene and aberrant methylation at the firstexon ofp16INK4a, were used to identify and characterize tumor and plasmaDNA. Thirteen clinicopathological parameters were analyzed in eachpatient. We identified 56 cases (90%) with at least one molecular event intumor DNA, and 41 cases (66%) with a similar alteration in plasma DNA.Comparison of the clinicopathological parameters between patients withand without plasma DNA revealed significant differences in the axillaryinvolvement, rate of invasive ductal carcinoma, high proliferative index,and the parameter comprised of lymph node metastases, histologicalgrade III, and peritumoral vessel involvement. A high proportion ofbreast cancer patients exhibited plasma DNA at diagnosis similar totumor DNA, and its presence correlated significantly with pathologicalparameters associated with a poor prognosis.

INTRODUCTION

Breast cancer is the most common malignancy in women in West-ern countries and accounts for 18.4% of all cancers in female patients(1). During recent decades, its prevalence rate has exhibited a clearincrease, in part as a result of the aging population; however, theage-specific prevalence has risen by 25% in the last 20 years, affecting1 in 10 women during their lifetime (2). We are also observing anincrease in breast cancer in young women under 35 years of age (3).Qualitative changes in the lifestyle of women in developed countriesthat can influence risk factors for breast cancer, such as age atmenarche, menopause, or first pregnancy, may partially explain thisphenomenon (2). In contrast, the mortality rate for breast cancer hasnot increased to the same degree as its prevalence because patientsurvival has improved due to new treatments. In addition to specifictreatments, the following strategies for reducing mortality should beconsidered: (a) a refinement of diagnostic techniques for early diag-nosis; and (b) the investigation of new methods able to detect meta-static or recurrent disease in preclinical or presymptomatic phases.

It is known that malignant cell transformation is accompanied bywell-defined molecular genetic changes within the original cell. To-

day, it is possible to detect many of them not only in research studiesbut also in clinical diagnostics in tumor DNA of tissue samplesobtained from diverse sources reached by the tumor cells, includingsputum (4), urine (5), pancreatic juice (6), and stool (7). In breastcancer, a case has been reported with c-erbB-2 gene amplificationfound in nipple discharge (8).

Previous studies have found that free DNA is circulating in bothhealthy and ill individuals. In control subjects, the mean concentrationof soluble DNA in plasma was 14 ng/ml (9), and in patients withdifferent types of neoplasias, the mean concentration rose to 180ng/ml (10). A mean concentration of 118 ng/ml was also detected inbenign gastrointestinal processes (9), and mean concentrations abovethose registered in controls have been reported in patients with auto-immune diseases such as systemic lupus erythematosus, rheumatoidarthritis, or other inflammatory conditions (11). An active processexplaining the enrichment of plasma DNA in cancer patients has alsobeen suggested (12). More recently, qualitative and quantitative stud-ies have demonstrated the presence of fetal DNA in maternal serum,with implications for noninvasive prenatal diagnosis (13–15). Like-wise, the examination of graft and host interactions has revealed thepresence of donor-specific DNA in the plasma of kidney and livertransplant recipients (16).

In recent years, using molecular techniques such as PCR for am-plification of small amounts of DNA, it has been possible to identifythe same alterations observed in tumor DNA in the plasma DNA ofpatients bearing diverse types of tumors. The alterations found includeK-ras, N-ras, and p53 gene mutations, aberrant promoter hyper-methylation of tumor suppressor genes, and changes in microsatellitesdetected by polymorphic markers in the following cancers: pancreaticcancer (17, 18); colon cancer (19–21); myelodysplastic syndromes(22); small cell lung cancer (23, 24); non-small cell lung cancer (25,26); head and neck carcinomas (27); clear cell renal carcinoma (28);and breast (24) and liver cancer (29).

Breast cancer is associated with different types of molecular geneticaberrations such as somatic mutations of oncogenes (30–32) andtumor suppressor genes (33–36) as well as allelic loss and MI3 inseveral chromosomal regions (37–42). The determination of theseanomalies can be used as a specific tool in the histological diagnosis,even early diagnosis, and can possibly be used as a prognostic factor.As observed in other tumors, the specific molecular alterations shownby breast carcinomas may also be found in the plasma DNA ofpatients harboring a breast cancer tumor.

Based on these facts, we designed the present study with two goals:(a) to determine the presence at diagnosis of tumor DNA in theplasma of patients with breast cancer, characterized by alterations inmicrosatellites and in tumor suppressor genes; and (b) to analyze thedistribution of 13 clinicopathological parameters in patients with andwithout specific plasma DNA.

Received 1/20/99; accepted 5/3/99.The costs of publication of this article were defrayed in part by the payment of page

charges. This article must therefore be hereby markedadvertisementin accordance with18 U.S.C. Section 1734 solely to indicate this fact.

1 Supported by grants from the Fundacion Caja Madrid and FIS number 98/0847.2 To whom requests for reprints should be addressed, at Molecular Genetics Unit,

Clinica del Trabajo, Avenida de la Reina Victoria 21, 28003-Madrid, Spain. Phone:34-1-533-8255; Fax: 34-1-554-9398; E-mail: felixbv@stnet.es.

3 The abbreviations used are: MI, microsatellite instability; dNTP, deoxynucleotidetriphosphate; LOH, loss of heterozygosity; CI, confidence interval; SSCP, single-strandconformational polymorphism; PVI, peritumoral vessel involvement; LNM, lymph nodemetastasis; IDC, invasive ductal carcinoma.

3251

Research. on February 5, 2014. © 1999 American Association for Cancercancerres.aacrjournals.org Downloaded from

PATIENTS AND METHODS

Tissue Sampling and DNA Extraction. All participants were informed ofthe nature of the study and gave their informed consent. Between October 1,1997 and October 30, 1998, samples from tumor tissue and correspondingnormal tissue were obtained sequentially, immediately after mastectomy, from62 patients with a diagnosis of breast cancer. They were then snap-frozen inliquid nitrogen until processing. All specimens underwent histological exam-ination to confirm the diagnosis of breast carcinoma. Pathological diagnosisand clinical evaluation disclosed no evidence of metastatic dissemination inany patient. A blood sample was collected from each patient with a clinicaldiagnosis of breast cancer on the day of surgery (before the mastectomy) toavoid the possible clearance of plasma DNA after removal of the primarytumor, and the sample was later discarded if histological diagnosis did notconclusively indicate the presence of a malignant lesion. Blood samples of 17healthy controls were also obtained. DNA was extracted from normal bloodcells and plasma immediately thereafter. DNA extraction from tumor andnormal tissue samples and from peripheral blood mononuclear cells (whichwere also used as normal DNA to avoid possible molecular alterations thatmight have occurred in normal breast tissues; Ref. 42) was performed by anonorganic method (Oncor, Gaithersburg, MD). Plasma DNA was purified onQiagen columns (Qiamp Blood Kit; Qiagen, Hilden, Germany) according tothe protocol for blood and body fluids, introducing the following modifica-tions. Between 7.5 and 12 ml of plasma were heated at 99°C for 5 min on aheat block. The heated sample was then centrifuged at 14,000 rpm for 30 min,after which the clear supernatant (about 1 ml) was collected (13). Proteinase K(20 mg/ml; Boehringer Mannheim, Mannheim, Germany) and buffer AL(Qiagen) were added in a 1:10 ratio with respect to the collected supernatantand incubated overnight at 55°C. One column was used repeatedly until thewhole sample had been processed. The DNA extracted was quantified spec-trophotometrically.

Analysis of Clinicopathological Parameters.The following parameterswere obtained from the medical records of the 62 patients studied: (a) birth anddiagnosis dates; (b) family history of the disease; (c) menopausal status; (d)tumor size; (e) LNMs; (f) presence of steroid receptors (estrogen and proges-terone); (g) histological type; (h) peritumoral vessel invasion; and (i) histo-logical grade. Pathological stage was assessed using the tumor-node-metastasis(TNM) classification. All tumors were graded with regard to two parameters:(a) nuclear polymorphism (uniform and regular size, 1; moderate pleomor-phism, 2; highly pleomorphic with giant nucleus, 3); and (b) mitotic index (,1mitosis, 1; 2, 2;.3, 3). The final grade was determined by adding the twoscores: grade 1, 1–2; grade 2, 3–4; grade 3, 5–6. Presence of peritumoralvessel invasion was analyzed, and the steroid receptor content (estrogen andprogesterone) was determined by an immunohistochemical procedure, theresults of which were considered to be positive when 25% or more of the cellsstained positively. The proliferative index of the tumors was demonstrated byKi-67 antigen (Immunotech, Westbrook, ME) in an immunohistochemicalanalysis; the Ki-67 labeling index was considered high when 15% or more ofthe cells stained positively.

Microsatellite Analysis and PCR Conditions. PCR was performed in25-ml volumes using 100 ng of template DNA, 0.75 unit of Ampli Taq GoldDNA polymerase (Perkin-Elmer, Roche Molecular Systems, Inc., Branchburg,NJ), 2.5-ml of 103 PCR buffer, 200mM dNTP, 0.6mM of each primer, anddifferent concentrations of MgCl2, depending on the polymorphic marker. A30-cycle amplification was done in a thermal cycler (Perkin-Elmer, FosterCity, CA). Six microsatellite markers were used to determine LOH on thefollowing chromosomes: (a) on chromosome 17, D17S855 (43) and D17S654(44); (b) on chromosome 16, D16S421 (45); (c) on chromosome 11, TH2 (42);(d) on chromosome 10, D10S197 (37); and (e) on chromosome 9, D9S161(46). These markers were chosen because they have been reported to show ahigh rate of alterations in breast carcinomas. The alleles were separated bymixing 25ml of the PCR products with a 10-ml volume of loading buffer (totalvolume, 35ml), 0.02% xylene cyanol, and 0.02% bromphenol blue. Electro-phoresis was run on nondenaturing 8–12% polyacrylamide gels for 12–15 h at500 V. After gel electrophoresis, the allelic band intensity was detected by anonradioisotopic technique using a commercially available silver stainingmethod (47). We analyzed the allelic intensities by densitometry. The gelimage was captured by a GS-690 Imaging Densitometer (Bio-Rad Laborato-ries, Hercules, CA), digitized at 400 dpi, and analyzed using Multi-Analyst/PC

(Bio-Rad Laboratories). An allele should be considered lost when its signal isreduced by 100% with respect to that observed in the normal counterpart DNA.However, it is accepted that possible variations in the amplification processand, more frequently, the presence of small amounts of normal cells among thetumor tissue or, in our study, of normal blood cells DNA in plasma DNA canresult in the expression of different band intensities in the total absence of oneallele. In this study, we considered LOH to exist when the signal of the allelewas reduced by more than 75% and considered MI to exist when one or morenovel bands appeared in tumor or plasma DNA. Despite the fact that the latteralteration is rare in breast carcinoma (37, 38), we checked for it among theresults of the microsatellite study.

Mutational Study of the p53 Gene. To establish the presence of pointmutations in the conserved exons of TP53, PCR-SSCP analysis was performedaccording to a modification of the method reported by Oritaet al. (48). Weamplified exons 5, 6, 7, and 8 of the p53 gene, and the primers used were: exon5, 59- TCCTTCCTCTTCCTACAG and 59-ACCCTGGGCAACCAGC-CCTGT; exon 6, 59-ACAGGGCTGGTTGCCCAGGGT and 59- AGTTG-CAAACCAGACCTCAGGCG; exon 7, 59-TCCTAGGTTGGCTCTGACTGTand 59-AGTGGCCCTGACCTGGAGTCT; and exon 8, 59-GGGACAGG-TAGGACCTGATTTCCTT and 59-ATCTGAAGGCATAACTGCACCCT-TGG. The annealing temperatures were 65°C, 67°C, 62°C and 68°C, respec-tively. PCR was performed under standard conditions in 25-ml that contained2-ml (100 ng) of DNA template (tumor, normal or plasma DNA); 2.5-ml of10 3 PCR buffer and 0.75 U of AmpliTaq Gold (Perkin-Elmer, RocheMolecular Systems Inc., Branchburg, NJ); 200mM dNTP mix; 0.6mM of eachprimer; and different concentrations of magnesium chloride depending on theprimer, and distilled H2O needed to reach the total volume. For PCR ampli-fication, the samples underwent 40 cycles of 94°C for 1 min, subjected todifferent annealing temperatures depending on the primer, and 70°C for 1 min.The amplified products were denatured by mixing with 15ml of denaturingstop solution that contained 98% formamide, 10 mml/L edathamil (pH 8.0),0.02% xylene cyanol and 0.02% bromphenol blue, heated to 95°C for 5 minand rapidly cooled on ice. Electrophoresis was run on nondenaturing 8%–12%polyacrylamide gels for 12–15 h at 250 v. The allelic band intensity on the gelswas detected by a nonradioisotopic method using a commercially availablesilver staining method (47). The specimens that showed a differential band atSSCP were amplified to obtain templates for DNA sequencing. These ampli-fications were independent from those used for SSCP analysis. AmplifiedDNA fragments were purified from 0.9% agarose gels using the Geneclean Kit(Bio-101, Inc. La Jolla, CA), and used for direct DNA sequencing by the dNTPmethod with the Sequenase Kit (United States Biochemical Corp. Cleveland,OH).

Methylation Study of the First Exon of p16INK4a. We also used themethylation status of the first exon ofp16INK4a as a molecular alteration toidentify patients with tumor DNA in plasma. The study was performed in the43 cases in which plasma DNA was available after microsatellite andp53genemutational analysis. A PCR-based methylation assay was performed, based onthe inability of some restriction enzymes to cut methylated sequences (49). ThePstI and SacII restriction enzyme sites were examined. Analysis of DNAdigests was performed according to the manufacturer’s instructions (Promega,Madison, WI), and the digested DNA was amplified with primers flanking therestriction sites. Amplification ofb-globin was used as internal control of thereaction in a multiplex PCR. (There are no restriction sites in theb-globinsequence selected forPstI andSacII.) The primer set used for methylationanalysis of exon 1 ofp16INK4a was 59-GGGAGCAGCATGGAGCCG and59-AGTCGCCCGCCATCCCCT, and the primer set forb-globin was 59-CAACTTCATCCACGTTCACC and 59-GAAGAGCCAAGGACAGGTAC.The conditions used were as follows: 2.5ml of 103 buffer II (Perkin-Elmer,Roche Molecular Systems Inc.); 200mM dNTPs; 2.5 mM MgCl2; 0.6 mM exon1 primers; 0.6mM b-globin primers; 100 ng of the digested DNA as template;and 25ml of distilled H2O. This reaction mix was amplified with Ampli TaqGold (Perkin-Elmer, Roche Molecular Systems Inc.) at 94°C for 12 min with35 cycles of 94°C for 30 s, 61°C for 30 s, and 72°C for 30 s, followed byincubation at 72°C for 11 min. PCR products were resolved in 6% acrylamideand stained with a nonisotopic silver nitrate method (47). PCR-based methy-lation analysis using restriction enzymes may be subject to variability if theDNA digestions are not complete. To rule out the possibility of incompleterestriction, all samples were digested overnight in two independent experi-

3252

PLASMA DNA IN BREAST CANCER PATIENTS

Research. on February 5, 2014. © 1999 American Association for Cancercancerres.aacrjournals.org Downloaded from

ments. PCR amplifications from each of the duplicate digests were repeatedtwice to ensure reproducibility of the results.

Statistical Analysis. A descriptive statistical study was performed in whichthe categorical variables were tabulated according to their absolute valueproportion and 95% CI. The continuous variables are expressed in terms of themean and 95% CI. The categorical variables were contrasted by means of thex2 test with the Yates correction (50) or Fisher’s exact test when any of theexpected frequencies was less than 5. We consideredP , 0.05 to be significantStatistical analyses were performed using the EPI-INFO package, version 6.04.

RESULTS

We studied 62 breast cancer patients for the presence of moleculargenetic alterations in their plasma and tumor DNA. The objective wasto detect molecular events in serum DNA that supported evidence ofits tumor origin. Thus, as molecular markers, we used the allelic statusof six different chromosomal regions (37, 42–46), point mutations ofthe p53 gene because they are detected in 16–81% of breast carci-nomas (51, 52), and the methylation status of the first exon ofp16INK4a, which has been reported in 31% of these tumors (53).

Plasma DNA was found in all 62 patients at concentrations rangingfrom 24–170 ng/ml (mean concentration, 115 ng/ml).

Microsatellite Analysis. Although the microsatellites used wereselected on the basis of their high rate of LOH in breast carcinomas,about 40% of our cases were found to be uninformative. Fifty-onebreast carcinomas (82%) showed allelic loss at at least one locus. Thesame analysis of plasma DNA disclosed 38 patients (61%) with thesame microsatellite alterations (Fig. 1). The highest individual rate ofLOH in tumor DNA corresponded to markers D17S654 and D16S421,with a 53% and 50% LOH, respectively (Table 1). All markers studiedshowed allelic loss in plasma DNA, but with variable percentageswith respect to tumor DNA; TH2 displayed the highest coincidencerate, with 100% of cases (Table 1). No MI was detected in tumorDNA. However, we observed that in five cases, plasma DNA dem-onstrated other additional molecular alterations not present in tumorDNA: two MI and one LOH for marker D17S654; and two LOH forD9S161.

Mutational Status of the p53 Gene. Tumor point mutations inTP53 were detected in seven tumors of our 62 patients (11%), anincidence that is lower than that reported elsewhere for breast cancer(Refs. 51 and 52; Figs. 2 and 3); in four of these tumors, LOH wasalso found in regions other than 17p13, and, in two tumors, LOH wasnot observed in any of the microsatellites studied. In three cases (5%),

the same mutation was identified in plasma DNA, including onepatient without concomitant LOH at any marker tested (Table 1).

Methylation of the First Exon of p16INK4a. We used this novelmolecular alteration (53) to identify cases in which DNA was presentin plasma. Ten cases were detected among 43 carcinomas studied(23%); three cases were detected in tumors with no allelic loss orp53gene mutations. Six patients (14%) exhibited aberrant methylation inplasma DNA (Fig. 4), two of them corresponded to cases with neitherLOH nor p53 gene mutations (Table 1).

In each molecular determination, microsatellite analysis,TP53SSCP study, and methylation assay, the DNA pattern of normal bloodcells and normal tissue was used as a control for comparison withtumor and plasma DNA. No differences were observed betweennormal blood cells and normal tissue DNA samples in every case.

In all, we identified 56 cases (90%) in which there was at least onemolecular event in tumor DNA and 41 cases (66%) with a similaralteration in plasma DNA. Of the 15 patients that showed molecularchanges in tumor DNA but not in plasma DNA, 9 patients displayedthe same molecular pattern in normal tissue, normal blood cells, andplasma DNA; in 6 cases, no amplification of the plasma DNA sampleswas obtained. The six cases in which there were no molecular alter-ations in the primary tumor had a mean plasma DNA concentration of70 ng/ml, but the molecular study revealed no differences amongnormal, normal blood cells, tumor, and plasma DNA.

Correlations between Clinicopathological Parameters and Mo-lecular Changes.To establish a clinical meaning for the presence ofplasma DNA with the characteristics of tumor DNA, we studied thecorrelation between the presence or absence of plasma DNA of tumororigin and 13 clinicopathological parameters at diagnosis (Table 2).Statistical analysis revealed statistically significant differences in the

Fig. 1. Representative photograph of two gels taken under normal light after stainingwith a (NO3)Ag method showing (A) microsatellite LOH at the D9S161 marker in patient25. Thearrow indicates the loss of the same allele in tumor and plasma and its presencein the normal tissue and lymphocyte DNA of the same patient.B, appearance of a newband in plasma DNA corresponding to MI at the D17S654 marker in case 12. Plasma,P;tumor,T; normal breast tissue,N; lymphocyte,L.

Fig. 2. SSCP photograph of exon 7 of thep53gene in case 14. Tumor (T) and plasma(P) presented extra bands (arrow) when compared with normal tissue (N) and lymphocyte(L) DNA from the same patient.

Table 1 Analysis of microsatellites, TP53 gene mutations, and aberrant methylation ofthe first exon of p16INK4a in tumor and plasma DNA of breast cancer patients

A. Marker n Ia Tumor LOH (%) Plasma LOH (%) P

TH2 62 35 6 (10) 6 (10) 0D10S197 62 30 9 (14) 6 (10) 0D16S421 62 22 11 (18) 8 (13) 0D17S855 62 36 13 (21) 8 (13) 0D17S654 62 38 20 (32) 11 (18) 3 (2 MI)D9S161 62 38 10 (16) 6 (10) 2

B. Tumor mutations (%) Plasma mutations (%)

TP53 62 7 (11) 3 (5) 0

C. Tumor methylation (%) Plasma methylation (%)

Methylation ofp16INK4a

43 10 (23) 6 (14) 0

a I, number of informative cases for each marker; P, number of cases with newalterations only in plasma DNA.

3253

PLASMA DNA IN BREAST CANCER PATIENTS

Research. on February 5, 2014. © 1999 American Association for Cancercancerres.aacrjournals.org Downloaded from

following independent variables: (a) involvement of three or morelymph nodes (P 5 0.005); (b) IDC (P 5 0.05); and (c) a highproliferative index (P 5 0.02). When we considered the subgroup ofpatients whose tumors concomitantly exhibited three characteristicsclassically associated with high-grade malignancy (LNMs, histologi-cal grade III, and PVI) and analyzed them with respect to the presenceor absence of plasma DNA, a significant difference (P 5 0.009) wasalso observed. The rest of the variables analyzed displayed no statis-tically significant differences (Table 2). During the 1-year studyperiod reported here, there were no cases of relapse or death, regard-less of the presence or absence of plasma DNA.

Plasma DNA of healthy controls was extracted in lower concen-trations than in cancer patients, ranging between 0 and 45 ng/ml. Thesame six markers were tested in plasma and normal blood cell DNAin the 13 cases in which the amount of plasma DNA extracted allowedthe performance of amplification. After microsatellite analysis, theavailable plasma DNA allowed the study ofTP53mutational status infive cases, and, after that, methylation of the first exon of p16INK4a

could also be performed in three of these cases. In each case, thesamples of plasma and normal blood cells examined displayed thesame DNA patterns.

DISCUSSION

The results of this study provide evidence of the presence of plasmaDNA with features of tumor DNA in breast cancer patients. In 1977,Leonet al. (10) reported the presence of free DNA at concentrationsranging between 0 and 2mg/ml in the serum of breast cancer patients

and reported that it was possible to analyze variations in the amountdepending on the stage of the disease and the response to the treatmentreceived by the patients; however, the specific characterization of thisDNA was not feasible. To date, several studies have reported thatcancer patients exhibit DNA in plasma, and molecular studies haveprovided evidence that this DNA is similar to tumor DNA (17–29).Overall, in our series, 66% of patients were found to present molec-ular alterations in plasma DNA. To optimize the results, it will beessential to obtain a panel of positive markers capable of detecting themolecular changes in plasma DNA for each tumor.

Another aim of this study was to determine whether the presenceof plasma DNA at diagnosis was significantly associated withsome of the clinicopathological characteristics of the patients.Among the 13 parameters evaluated, we found 4 that showedstatistically significant differences between patients with and with-out plasma DNA with tumor characteristics: (a) three (high pro-liferating index, three or more affected lymph nodes, and IDC) asindependent variables; and (b) one comprised of PVI, axillaryinvolvement, and high histological grade, which are three classicalpathological parameters of poor prognosis (Table 2). Our patientsshowed no significant differences in other independent variablessuch as tumor size, stage, histological grade, steroid receptors, orPVI that are also indicative of tumor extension and aggressivenessand, secondarily, of tumor cell turnover and capacity for metastaticspread. To date, six studies have analyzed the presence of tumorDNA in plasma and disease stage, and the results have not beenuniform. Two studies in head and neck carcinoma (27) and smallcell lung cancer (23) reported a large number of patients withadvanced disease and the presence of tumor DNA in plasma, andfour studies also associated this molecular event with the earlystages of non-small cell lung cancer (25), colon cancer (19, 20),and clear cell renal carcinoma (28). Taken together, the dataavailable concerning the extension of tumor disease and the ap-pearance of plasma DNA with tumor characteristics do not offer aconclusive explanation, suggesting that other unexplored biologi-cal characteristics of the tumor cells could be related to thisphenomenon. Thus, the results of our study may indicate that thepresence of plasma DNA of tumor origin in breast cancer patientsis significantly associated with some histological features of highlymalignant lesions.

In addition to the possible role of free plasma DNA as a prognosticfactor in breast cancer, other applications can be also considered. Theearly diagnosis of breast cancer is currently one of the best strategiesfor improving the survival of these patients. If the results of thisprocedure were found to be positive in disease stage I or with tumorsmeasuring less than 1 cm, it would be a good implementation of the

Fig. 3. Sequencing analysis in case 14 shows a nucleotide change (C to T) in thereverse sequencing that corresponds to a transition (G to A) in exon 7 of thep53 gene.This transition causes an amino acid change (Arg-Trp).

Fig. 4. Photograph of a gel showing the PCR-based methylation assay of exon 1 ofp16INK4a in case 20.C, lymphocyte DNA of a control case displaying amplified productsfor b-globin and exon 1 ofp16INK4a. After digestion with restriction enzymePstI, tumorDNA (T) and plasma DNA (P) amplified for exon 1 ofp16INK4a indicate the presence ofmethylation. In contrast, lymphocyte DNA (L) and normal breast tissue DNA (N)exhibited no amplification of exon 1 ofp16INK4a, suggesting the absence of DNAmethylation.b-Globinwas used as an internal control because none of the restriction sitesselected forPstI andSacII exist in its sequence.

Table 2 Pathological characteristics of breast carcinomas showing a statisticallysignificant difference between patients presenting plasma DNA with and without

tumor features

Tumor plasma DNA No tumor plasma DNA

PN % 95% CI N %

No. of eligible patients 41 15LNMs

,3 26 63 (47–78) 15 100 (78–100) 0.005$3 15 37 (22–53) 0 0 (0–22)

HistologyIDC 33 80 (65–91) 8 53 (27–79) 0.05Others 8 20 (9–35) 7 47 (21–73)

PVI, LNM, grade IIIYes 13 32 (18–48) 0 0 (0–22) 0.009No 28 68 (52–82) 15 100 (78–100)

Ki-67 indexHigh 32 78 (62–89) 7 47 (21–73) 0.02Low 9 22 (11–38) 8 53 (27–79)

3254

PLASMA DNA IN BREAST CANCER PATIENTS

Research. on February 5, 2014. © 1999 American Association for Cancercancerres.aacrjournals.org Downloaded from

current radiographic procedures. Moreover, the identification of pre-malignant disease with defined molecular alterations (54) would per-mit the investigation of these molecular events in patients with pre-neoplastic lesions, hypothetically changing the management of theseprocesses.

Previous observations have suggested that plasma DNA can un-dergo quantitative changes in cancer patients after radiation therapy(10); however, its utility in monitoring chemotherapy and radiother-apy has yet to be fully defined.

The identification of recurrent breast cancer at a preclinicalstage is an essential strategy in the fight against this disease, but noreliable serological predictive markers are currently available (55).Considering the prevalence and incidence of this disease, thediscovery of a valid marker could have a strong impact on itsmanagement and probably on survival. We consider that the utili-zation of a simpler methodology that, as in our case, does notinvolve radioisotopic techniques may help to introduce plasmaDNA analysis into the study and management of breast cancer. Inthe coming years, prospective studies, such as that recently re-ported in pancreatic carcinomas (56), will provide us with defini-tive information on the value of this research tool as a prognosticfactor in cancer patients.

ACKNOWLEDGMENTS

We are indebted to the women who participated in the study; to M. Rosado,Z. Garcia, and M. C. Manzanares for help with the collection of blood samples;to M. Messman for revision and preparation of the manuscript; and to I. Millanfor the statistical study.

REFERENCES

1. Kelsey, J. L., and Horm-Ross, P. L. Breast cancer: magnitude of the problem anddescriptive epidemiology. Epidemiol. Rev.,15: 7–16, 1993.

2. Veronesi, U., Goldhirsch, A., and Yarnold, J. Breast cancer.In: M. Peckham, H. M.Pinedo, and U. Veronesi (eds.), Oxford Textbook of Oncology, Vol. 2, pp. 1243–1289. New York, NY: Oxford University Press, 1995.

3. Lopez-Abente, G., Pollan, M., Escolar, A., Errezola, M., and Abraira, V. Atlas ofCancer Mortality and Other Causes of Death in Spain, 1978–1992. Madrid: Funda-cion Cientıfica de la Asociacion Espanola contra el Cancer, 1996.

4. Mao, L., Hruban, R. H., Boyle, J. O., Tockman, M., and Sidransky, D. Detection ofoncogene mutations in sputum precedes diagnosis of lung cancer. Cancer Res.,54:1634–1637, 1994.

5. Sidransky, D., Von Eschenbach, A., Tsai, Y. C., Jones, P., Summerhayes, I., Mar-shall, F., Paul, M., Green, P., Hamilton, S. R., Frost, P., and Vogelstein, B. Identi-fication of p53 gene mutations in bladder cancers and urine samples. Science(Washington DC),252: 706–709, 1991.

6. Kondo, H., Sugano, K., Fukamaya, N., Kyogoku, A., Nose, H., Shimada, K., Ohkura,H., Ohtsu, A., Yishida, S., and Shimosato, Y. Detection of point mutations in theK-ras oncogene at codon 12 in pure pancreatic juice for diagnosis of pancreaticcarcinoma. Cancer (Phila.),73: 1589–1594, 1994.

7. Sidransky, D., Tokino, T., Hamilton, S. R., Kinzler, K. W., Levin, B., Frost, P., andVogelstein, B. Identification ofras oncogene mutations in the stool of patients withcurable colorectal tumors. Science (Washington DC),256: 102–105, 1992.

8. Motomura, K., Koyama, H., Noguchi, S., Inaji, H., and Azuma, C. Detection ofc-erbB-2 gene amplification in nipple discharge by means of polymerase chainreaction. Breast Cancer Res. Treat.,33: 89–92, 1995.

9. Shapiro, B., Chakrabarty, M., Cohn, E. M., and Leon, S. A. Determination ofcirculating DNA levels in patients with benign or malignant gastrointestinal disease.Cancer (Phila.),51: 2116–2120, 1983.

10. Leon, S. A., Shapiro, B., Sklaroff, D. M., and Yaros, J. Free DNA in the serum ofcancer patients and the effect of therapy. Cancer Res.,37: 646–650, 1977.

11. Koffler, D., Agnello, V., Winchester, R., and Kunkel, H. G. The occurrence ofsingle-stranded DNA in the serum of patients with SLE and other diseases. J. Clin.Investig.,52: 198–204, 1973.

12. Stroun, M., Anker, P., Maurice, P., Lyautey, J., Lederrey, C., and Beljanski, M.Neoplastic characteristics of the DNA found in the plasma of cancer patients.Oncology,46: 318–322, 1989.

13. Lo, Y. M. D., Corbetta, N., Chamberlain, P. F., Rai, V., Sragent, I. L., Redman,C. W. G., and Wainscoat, J. S. Presence of fetal DNA in maternal plasma and serum.Lancet,350: 485–487, 1997.

14. Lo, Y. M., Tein, M. S., Lau, T. K., Haines, C. J., Leung, T. N., Poon, P. M.,Wainscoat, J. S., Johnson, P. J., Chang, A. M., and Hjelm, N. M. Quantitative analysisof fetal DNA in maternal plasma and serum: implications for noninvasive prenataldiagnosis. Am. J. Hum. Genet.,62: 768–775, 1998.

15. Lo, Y. M. D., Hjelm, N. M., Fidler, C., Sargent, I. L., Murphy, M. F., Chamberlain,P. F., Poon, P. M. K., Redman, C. W. G., and Wainscoat, J. S. Prenatal diagnosis offetal RhD status by molecular analysis of maternal plasma. N. Engl. J. Med.,339:1734–1738, 1998.

16. Lo, Y. M. D., Tein, M. S. C., Pang, C. C. P., Yeung, C. K., Tong, K-L., and Hjelm,N. M. Presence of donor-specific DNA in plasma of kidney and liver-transplantrecipients. Lancet,351: 1329–1330, 1998.

17. Mulcahy, H. E., Lyautey, J., Lederrey, C., Cheng, X. Q., Anker, P., Alstead, E. M.,Ballinger, A., Farthing, M. J., and Stroun, M. A prospective study of K-ras mutationsin the plasma of pancreatic cancer patients. Clin. Cancer Res.,4: 271–275, 1998.

18. Sorenson, G. D., Pribish, D. M., Valone, F. H., Memoli, V. A., Bzik, D. J., and Yao,S. L. Soluble normal and mutated DNA sequences from single-copy genes in humanblood. Cancer Epidemiol. Biomark. Prev.,3: 67–71, 1994.

19. Anker, P., Lefort, F., Vasioukhin, V., Lyautey, J., Lederrey, C., Cheng, X. Q., Stroun,M., Mulcahy, H. E., and Farthing, J. G. K-ras mutations are found in DNA extractedfrom the plasma of patients with colorectal cancer. Gastroenterology,112: 1114–1120, 1997.

20. Hibi, K., Robinson, C. R., Booker, S., Wu, L., Hamilton, S. R., Sidransky, D., andJen, J. Molecular detection of genetic alterations in serum of colorectal cancerpatients. Cancer Res.,58: 1405–1407, 1998.

21. Kopreski, M. S., Benko, F. A., Kwee, C., Leitzel, K. E., Eskander, E., Lipton, A., andGocke, C. D. Detection of mutant K-ras DNA in plasma or serum of patients withcolorectal cancer. Br. J. Cancer,76: 1293–1299, 1997.

22. Vasioukhin, V., Anker, P., Maurice, P., Lyautey, J., Lederrey, C., and Stroun, M.Point mutations of the N-ras gene in the blood plasma DNA of patients withmyelodysplastic syndrome or acute myelogenous leukaemia. Br. J. Haematol.,86:774–779, 1994.

23. Chen, X. Q., Stroun, M., Magnenat, J. L., Nicod, N. P., Kurt, A-M., Lyautey, J.,Lederrey, C., and Anker, P. Microsatellite alterations in plasma DNA of small celllung cancer patients. Nat. Med.,2: 1033–1035, 1996.

24. Silva, J. M., Gonzalez, R., Dominguez, G., Garcia, J. M., Espana, P., and Bonilla, F.TP53 gene mutations in plasma DNA of cancer patients. Genes ChromosomesCancer,24: 160–161, 1999.

25. Sanchez.-Cespedes, M., Monzo, M., Rosell, R., Pifarre, A., Calvo, R., Lopez-Cabrerizo, M. P., and Astudillo, J. Detection of chromosome 3p alterations in serumDNA of non-small-cell lung cancer patients. Ann. Oncol.,9: 113–116, 1998.

26. Esteller, M., Sanchez-Cespedes, M., Rosell, R., Sidransky, D., Baylin, S. B., andHerman, J. G. Detection of aberrant promoter hypermethylation of tumor suppressorgenes in serum DNA from non-small cell lung cancer patients. Cancer Res.,59:67–70, 1999.

27. Nawroz, H., Koch, W., Anker, P., Stroun, M., and Sidransky, D. Microsatellitealterations in serum DNA of head and neck cancer patients. Nat. Med.,2: 1035–1037,1996.

28. Goessl, C., Heicapell, R., Munker, R., Anker, P., Stroun, M., Krause, H., Muller, M.,and Miller, K. Microsatellite analysis of plasma DNA from patients with clear cellrenal carcinoma. Cancer Res.,58: 4728–4732, 1998.

29. Wong, I. H. N., Lo, Y. M. D., Zhang, J., Liew, C-T., Ng, M. H. L., Wong, N., Lai,P. B. S., Lau, W. Y., Hjelm, N. M., and Johnson, P. J. Detection of aberrant p16methylation in the plasma and serum of liver cancer patients. Cancer Res.,59: 71–73,1999.

30. Theillet, C., Lidereau, R., Escot, C., Hutzell, P., Brunet, M., Gest, J., Schlom, J., andCallahan, R. Loss of a c-H-ras-1 allele and aggressive human primary breast carci-nomas. Cancer Res.,46: 4776–4781, 1986.

31. Mangues, R., Kahn, J. M., Seidman, I., and Pellicer, A. An overexpressed N-rasproto-oncogene cooperates withN-methylnitrosourea in mouse mammary carcino-genesis. Cancer Res.,54: 6395–6401, 1994.

32. Kumar, R., Sukumar, S., and Barbacid, M. Activation ofrasoncogenes preceding theonset of neoplasia. Science (Washington DC),248: 1101–1104, 1990.

33. Crook, T., Crossland, S., Crompton, M. R., Osin, P., and Gusterson, B. A.p53mutations in BRCA1-associated familial breast cancer. Lancet,350: 638 – 639,1997.

34. Sjogren, S., Inganas, M., Norberg, T., Lindgren, A., Nordgren, H., Holmberg, L.,and Bergh, J. Thep53 gene in breast cancer: prognostic value of complementaryDNA sequencingversusimmunohistochemistry. J. Natl. Cancer Inst.,88: 173–182, 1996.

35. Hollstein, M. Sidransky, D., Vogelstein, B., and Harris, C. C.p53mutations in humancancers. Science (Washington DC),253: 49–53, 1991.

36. Gretarsdottir, S., Thorlacius, S., Valgardsdottir, R., Gudlaugsdottir, S., Sigurdsson, S.,Steinarsdottir, M., Jonasson, J. G., Anamthawat-Jonsson, K., and Eyfjord, J. E.BRCA2andp53mutations in primary breast cancer in relation to genetic instability.Cancer Res.,58: 859–862, 1998.

37. Patel, U., Grundfest-Broniatowski, S., Gupta, M., and Banerjee, S. Microsatelliteinstabilities at five chromosomes in primary breast tumors. Oncogene,9: 3695–3700,1994.

38. Toyama, T., Iwase, M., Yamashita, H., Iwata, H., Yamashita, T., Ito, K., Hara, Y.,Suchi, M., Kato, T., Nakamura, T., and Kobayashi, S. Microsatellite instability insporadic human breast cancer. Int. J. Cancer,68: 447–451, 1996.

39. Phelan, C. M., Borg, A., Cuny, M., Crichon, C. N., Baldersson, T., Andersen, T. I.,Caligo, M. A., Lidereau, R., Lindblom, A., Seitz, S., Kelsell, D., Hamann, U., Rio, P.,Thorlacius, S., Papp, J., Olah, E., Ponder, B., Bignon, Y-J., Scherneck, S., Barkar-dottir, R., Borresen-Dale, A-L., Eyfjord, J., Theillet, C., Thompson, A. M., Devilee,P., and Larsson, C. Consortium study on 1280 breast carcinomas: allelic loss onchromosome 17 target subregions associated with family history and clinical param-eters. Cancer Res.,58: 1004–1012, 1998.

40. Hamann, U., Herbold, C., Costa, S., Solomayer, E-F., Kaufmann, M., Bastert, G.,Ulmer, H. U., Frenzel, H., and Komitowski, D. Allelic imbalance on chromosome

3255

PLASMA DNA IN BREAST CANCER PATIENTS

Research. on February 5, 2014. © 1999 American Association for Cancercancerres.aacrjournals.org Downloaded from

13q: evidence for the involvement of BRCA2 and RB1 in sporadic breast cancer.Cancer Res.,56: 1988–1990, 1996.

41. Noviello, C., Courjal, F., and Theillet, C. Loss of heterozygosity on the long arm ofchromosome 6 in breast cancer: possibly four regions of deletions. Clin. Cancer Res.,2: 1601–1606, 1996.

42. Deng, G., Lu, Y., Zlotnikov, G., Thor, A. N., and Smith, H. S. Loss of heterozygosityin normal tissue adjacent to breast carcinomas. Science (Washington DC),274:2057–2059, 1996.

43. Anderson, L. A., Friedman, L., Osborne-Lawrence, S., Lynch, E., Weissenbach, J.,Bowcock, A., and King, M-C. High-density genetic map of the BRCA1 region ofchromosome 17q12–q21. Genomics,17: 618–623, 1993.

44. Steiner, G., Schoenberg, M. P., Linn, J. F., Mao, L., and Sidransky, D. Detection ofbladder cancer recurrence by microsatellite analysis of urine. Nat. Med.,3: 621–624,1997.

45. Skirnisdottir, S., Eiriksdottir, G., Baldersson, T., Barkardottir, R. B., Egilsson, V., andIngvarsson, S. High frequency of allelic imbalance at chromosome region 16q22–23in human breast cancer: correlation with high PgR and low S phase. Int. J. Cancer,64:112–116, 1995.

46. Brenner, A. J., and Aldaz, C. M. Chromosome 9p allelic loss and p16/CDKN2 inbreast cancer and evidence of p16 inactivation in immortal breast epithelial cells.Cancer Res.,55: 2892–2895, 1995.

47. Oto, M., Miyake, S., and Yuasa, Y. Optimization of nonradioisotopic single strandconformation polymorphism analysis with a conventional minislab gel electrophore-sis apparatus. Anal. Biochem.,213: 19–22, 1993.

48. Orita, M., Suzuki, Y., Sekiya, T., and Hayashi, K. Rapid and sensitive detection ofpoint mutations and DNA polymorphisms using the polymerase chain reaction.Genomics,5: 874–879, 1989.

49. Singer-Sam, J., Yang, T. P., Mori, N., Tanguay, R. L., Le-Bon, J. M., Flores, C. J.,and Riggs, A. D. DNA methylation in the 59 region of the mousePGK-1gene and a

quantitative PCR assay for methylation.In: G. A. Clawson, D. B. Willis, A. Weiss-bach, and P. A. Jones (eds.), Nucleic Acid Methylation, Vol. 128, pp. 285–289. NewYork: Alan Bliss, Inc., 1990.

50. Armitage, P., and Berry, G. Statistical Methods in Medical Research. Oxford:Blackwell Scientific Publications, 1987.

51. Norberg, T., Lennerstrand, J., Inganas, M., and Bergh, J. Comparison between p53protein measurements using the luminometric immunoassay and immunohistochem-istry with detection ofp53 gene mutations using cDNA sequencing in human breasttumors. Int. J. Cancer,79: 376–383, 1998.

52. Blaszyk, H., Hartmann, A., Tamura, Y., Saitoh, S., Cunningham, J. M., Mcgovern,R. M., Schroeder, J. J., Schaid, D. J., Ii, K., Monden, Y., Morimoto, T., Komaki, K.,Sasa, M., Hirata, K., Okazaki, M., Kovach, J. S., and Sommer, S. S. Molecularepidemiology of breast cancer in northern and southern Japan: the frequency, clus-tering, and patterns ofp53 gene mutations differ among these two low-risk popula-tions. Oncogene,13: 2159–2166, 1996.

53. Herman, J. G., Merlo, A., Mao, L., Lapidus, R. G., Issa, J-PJ., Davidson, N. E.,Sidransky, D., and Baylin, S. Inactivation of theCDKN2/p16/MTS1gene is frequentlyassociated with aberrant DNA methylation in all common human cancers. CancerRes.,55: 4525–4530, 1995.

54. O’Connell, P., Pekkel, V., Fuqua, S. A. W., Osborne, C. K., Clark, G. M., and Allred,D. C. Analysis of loss of heterozygosity in 399 premalignant breast lesions at 15genetic loci. J. Natl. Cancer Inst.,90: 697–703, 1998.

55. Caponigro, F. CA 15.3 in human breast cancer. Comparison with tissue polypeptideantigen (TPA) and carcinoembryonic antigen (CEA). Int. J. Biol. Markers,5: 73–76,1990.

56. Castells, A., Puig, P., Mora, J., Boadas, J., Boix, L., Urgell, E., Sole, M., Capella, G.,Lluis, F., Fernandez-Cruz, L., Navarro, S., and Farre´ A. K-ras mutations in DNAextracted from the plasma of patients with pancreatic carcinoma: diagnostic utilityand prognostic significance. J. Clin. Oncol.,17: 578–584, 1999.

3256

PLASMA DNA IN BREAST CANCER PATIENTS

Research. on February 5, 2014. © 1999 American Association for Cancercancerres.aacrjournals.org Downloaded from