VOLUME 24

NUMBER 26

SEPTEMBER 10 2006

JOURNAL OF CLINICAL ONCOLOGY

O R I G I N A L

R E P O R T

Prediction of Breast Tumor Progression by Integrity of Free Circulating DNA in Serum

Naoyuki Umetani, Armando E. Giuliano, Suzanne H. Hiramatsu, Farin Amersi, Taku Nakagawa, Silvana Martino, and Dave S.B. Hoon
From the Department of Molecular Oncology and the Joyce Eisenberg Breast Center, John Wayne Cancer Institute; and The Angeles Clinic and Research Institute, Santa Monica, CA. Submitted January 28, 2006; accepted May 19, 2006. Presented in part at the fourth International Conference on Circulating Nucleic Acids in Plasma/Serum, London, United Kingdom, September 4-6, 2005. Supported in part by the Komen Breast Cancer Foundation, California Breast Cancer Research Program, and Avon Breast Foundation. Terms in blue are dened in the glossary, found at the end of this article and online at www.jco.org. Authors disclosures of potential conicts of interest and author contributions are found at the end of this article. Address reprint requests to Dave S.B. Hoon, PhD, Department of Molecular Oncology, John Wayne Cancer Institute, 2200 Santa Monica Blvd, Santa Monica, CA 90404; e-mail: hoon@jwci.org. 2006 by American Society of Clinical Oncology 0732-183X/06/2426-4270/$20.00 DOI: 10.1200/JCO.2006.05.9493

Purpose Cell-free DNA circulating in serum is a candidate molecular biomarker for malignant tumors. Unlike uniformly truncated DNA released from apoptotic cells, DNA released from dead cancer cells varies in size. Serum DNA integrity, the ratio of longer fragments to total DNA, may be clinically useful for detecting breast cancer progression. Patients and Methods Serum samples from 51 healthy females and 83 females with primary breast cancers (eight American Joint Committee on Cancer stage 0, 24 stage I, 27 stage II, 21 stage III, and three stage IV) were assessed preoperatively. Serum DNA integrity was assessed by fragment lengthdependent quantitative real-time polymerase chain reaction of ALU DNA repeats. Results Mean serum DNA integrity was signicantly higher in patients with stage II, III, and IV breast cancers than in healthy females (P .005, P .0001, and P .002, respectively). The receiver operating characteristic (ROC) curve for discriminating patients with stage II or more advanced breast cancers from healthy females had an area under the curve (AUC) of 0.79 (95% CI, 0.70 to 0.86). Mean serum DNA integrity was positively correlated to size of invasive cancers (r 0.48; P .0001) and signicantly higher in the presence of lymphovascular invasion (LVI; 0.25 0.02 v 0.17 0.02; P .0001) or lymph node (LN) metastasis (0.27 0.02 v 0.14 0.02; P .0001). The ROC curve for discriminating LN metastasis had an AUC of 0.81 (95% CI, 0.72 to 0.89). Serum DNA integrity and LVI were signicant for predicting LN metastasis in a multivariate analysis (P .0002 and P .0001, respectively). Conclusion Integrity of serum circulating DNA is a promising molecular biomarker for detecting breast cancer tumor progression and regional LN metastases. J Clin Oncol 24:4270-4276. 2006 by American Society of Clinical Oncology

INTRODUCTION

Breast cancer is the second leading cause of cancer death among women in the United States, and accounted for one third of all new cancer cases among women in 2005.1 The most signicant prognostic indicator for patients with breast cancer is axillary lymph node (LN) metastasis.2 Physical examination and diagnostic imaging methods, such as ultrasonography or computed tomography, are effective for detection of LN metastases when they are relatively large. However, there is no clinically established blood test which has the predictive ability to determine regional LN or distant metastasis. Therefore, the development of a preoperative blood test for detection of LN metastasis of breast cancer is clinically desired.
4270

Tumor-related cell-free DNA circulating in blood is a promising candidate biomarker for malignant tumor detection or prognosis.3-6 Absolute levels of circulating DNA detected in serum/plasma have been related to presence7 and prognosis8 of breast cancer. Methylation of tumor suppressor genes detected in circulating DNA has demonstrated prognostic potential.9-12 Recently, it was shown that integrity of circulating DNA, measured as the ratio of longer to shorter DNA fragments, is higher in patients with gynecologic and breast cancers than in normal individuals.13 Apoptotic cells release DNA fragments that are usually 185 to 200 base pair (bp) in length14; this uniformly truncated DNA is produced by a programmed enzymatic cleavage process during apoptosis.15 In healthy individuals, the main source of free-circulating DNA is apoptotic cells. In contrast, DNA released from

Downloaded from jco.ascopubs.org on November 24, 2011. For personal use only. No other uses without permission. Copyright 2006 American Society of Clinical Oncology. All rights reserved.

Serum DNA Integrity in Breast Cancer Patients

malignant cells varies in size because pathologic cell death in the malignant tumors results not only from apoptosis, but also necrosis, autophagy, or mitotic catastrophe.16 Therefore, elevated levels of long DNA fragments may be a good marker for detection of malignant tumor DNA in blood.17 However, quantication of free circulating DNA in blood has not been practically utilized because of the difculty in handling the minute amount of DNA. Recently, we have developed a simple, robust, highly sensitive, and high throughput method to directly measure the serum DNA integrity. Serum was considered a better source of circulating DNA than plasma because serum contains signicantly higher amount of DNA with a low level of contaminating extraneous DNA released from leukocytes.18 In this assay, quantitative real-time PCR (qPCR) dependent on DNA fragment size was used to quantify the DNA in 0.1 L of serum using the ALU repeated DNA sequence. ALU repeats are the most abundant sequences in the human genome, with a copy number of about 1.4 million per genome.19,20 ALU sequences are short interspersed elements (SINEs), typically 300 nucleotides, which account for more than 10% of the genome.21 In this assay, purication of DNA from serum was unnecessary. DNA integrity was calculated as a ratio of longer to shorter DNA fragments, quantied by qPCR (115bp and 247bp) for ALU repeats. Applying this approach, we demonstrated that assessing serum DNA integrity was useful for preoperative prediction of regional LN metastasis in breast cancer. We also demonstrated that serum DNA integrity directly correlated with advancing American Joint Committee on Cancer (AJCC) breast cancer staging.
PATIENTS AND METHODS
Serum Samples and Clinical and Pathology Information Serum samples from 51 healthy females and 83 preoperative females with AJCC stage 0 to IV primary breast cancers (eight stage 0, 24 stage I, 27 stage II, 21 stage III, and three stage IV) were assessed. Blood was drawn before surgery of primary breast cancer. Staging was based on postoperative histopathology ndings for stages 0 to III, and imaging diagnoses were utilized for stage IV. All patients were selected by the database coordinator from the patients seen between 2000 and 2005 at the Joyce Eisenberg-Keefer Breast Center and the Angeles Clinic at Saint Johns Health Center, Santa Monica, CA. All patients in this study gave consent according to guidelines set forth by Saint Johns Health Center institutional review board. Serum Preparation for Direct qPCR Ten ml of blood was collected in a CORVAC serum separator tube (Sherwood-Davis & Geck, St Louis, MO) containing clot activation additive and barrier gel, stored at 4C, and processed within 6 hours: the blood was separated by centrifugation (1,000 g, 15 minutes) and passed through a 13-mm serum lter (Fisher Scientic, Pittsburgh, PA) to remove potentially contaminating cells. The serum was immediately cryopreserved at 80C. To deactivate or eliminate proteins binding to template DNA or DNA polymerase that might invalidate qPCR results, 20 L of each serum sample was mixed with 20 L of a preparation buffer that contained 2.5% of tween-20, 50 mmol/L Tris, and 1 mmol/L EDTA. This was digested with 16 g of proteinase K solution (Qiagen, Valencia, CA) at 50C, followed by 5 minutes of heat deactivation and insolubilization at 95C. After subsequent centrifugation of 10,000 g for 5 minutes, 0.2 L of supernatant was used as a template for each qPCR reaction. Quantitative PCR of ALU Repeats To achieve the highest sensitivity for DNA quantication, we applied a novel qPCR assay that utilizes primer sets to amplify the consensus ALU
www.jco.org

sequence (Fig 1A). Two sets of ALU primers were designed: the primer set for the 115bp amplicon (ALU115) amplies both shorter (truncated by apoptosis) and longer DNA fragments, whereas the primer set for the 247bp amplicon (ALU247) amplies only longer DNA fragments (Fig 1B). The sequences of the ALU115 primers were forward: 5 -CCTGAGGTCAGGAGTTCGAG-3 and reverse: 5 -CCCGAGTAGCTGGGATTACA-3 ; ALU247 primers were forward: 5 -GTGGCTCACGCCTGTAATC-3 and reverse: 5 -CAGGCTG GAGTGCAGTGG-3 . ALU115-qPCR results represent the total amount of free serum DNA; the ALU115 primer can amplify most fractions of circulating DNA. ALU247-qPCR results represent amounts of DNA released from nonapoptotic cells. DNA integrity was calculated as the ratio of qPCR results (ALU247-qPCR/ALU115-qPCR). Because the annealing sites of ALU115 are within the ALU247 annealing sites, the qPCR ratio (DNA integrity) is 1.0 when the template DNA is not truncated and 0.0 when all template DNA is completely truncated into fragments smaller than 247bp. The reaction mixture for each direct qPCR consisted of a template sequence, 0.2 mol/L of the forward and reverse primers (ALU115 or ALU247), one unit of iTaq DNA polymerase (Bio-Rad, Hercules, CA), 0.02 L of uorescein calibration dye (Bio-Rad), and 1 concentration of SYBR Gold (Molecular Probe, Eugene, OR) in a total reaction volume of 20 L with 5 mmol/L Mg2 . Real-time PCR amplication was performed with precycling heat activation of DNA polymerase at 95C for 10 minutes, followed 35 cycles of denaturation at 95C for 30s, annealing at 64C for 30s, and extension at 72C for 30s using iCycler iQ Real-Time PCR (Bio-Rad). The absolute equivalent amount of DNA in each sample was determined by a standard curve with serial dilutions (10ng-0.01pg) of gently prepared genomic DNA obtained from peripheral blood leukocytes of healthy donor volunteers. A negative control (without template) was performed in each plate. All qPCR assays were performed in a blinded fashion without knowledge of specimen identity. Mean values were calculated from triplicate reactions. Statistical Analysis Mean values for healthy females and patients with each stage of breast cancer were compared using Dunnetts multiple comparison. Mean values between two groups were compared using Students t test. Spearmans coefcients were used to identify the trends between AJCC stage and serum DNA integrity. Clinicopathology effect and serum DNA integrity in presence of LN metastasis was assessed using a nominal logistic regression model for multivariate analysis; variables indicated by the univariate analyses (P .10) were entered. Linear regression and multiple linear regression was used to assess the possible dependence of mean serum DNA integrity on patient age and tumor clinicopathology information. SAS version 5.0.1 (SAS Institute, Cary, NC) was used to conduct statistical analyses and MedCalc version 8.0 (MedCalc Software, Mariakerke, Belgium) was used for receiver operating characteristic (ROC) curve analyses. A P value less than .05 (two tailed) was considered signicant.

RESULTS

Clinical and Pathology of Primary Breast Cancers The mean age was 45 14 (standard deviation [SD]) years for 51 healthy females and 58 12 years for 83 patients with primary breast cancer. Table 1 shows the AJCC stage and histopathology characteristics of patients whose sera were sampled preoperatively. All the stage 0 cancers were ductal carcinomas in situ; the mean patient age was 59 11 (SD) years and the mean tumor size was 2.5 1.9 (SD) cm. Of the 42 patients with regional LN metastases, 10 had solitary micrometastases (pN1mi 2mm in size). Circulating DNA in Serum and AJCC Stage of Primary Breast Cancers Circulating DNA in serum of patients blood taken preoperatively was assessed for levels and integrity of serum DNA. An ALU115qPCR value represents the absolute total amount of serum DNA.
4271

Downloaded from jco.ascopubs.org on November 24, 2011. For personal use only. No other uses without permission. Copyright 2006 American Society of Clinical Oncology. All rights reserved.

Umetani et al

Fig 1. (A) A consensus of human longform ALU interspersed repetitive sequence with indication of the primers by : ALU115 primers (115 open boxes. bp amplicon); solid underlines ___ : ALU247 primers (247 bp amplicon). (B) Calculated relative efciency of DNA quantication in terms of DNA length. Estimated efciency of ALU115 primers is shown as a solid line: 0% in less than 115 bp and 90% at 1150 bp; ALU247 primers as a dotted line: 0% in less than 247 bp and 90% at 2470 bp. Dotted area represents the size of DNA released from apoptotic cells.

Because the absolute serum DNA levels tted exponential distribution, we applied logarithmic transformation to each value. The mean logarithmically transformed ALU115-qPCR values in healthy females and patients with stage 0, I, II, III, and IV breast cancer were 2.43 0.10 (SEM), 2.88 0.16, 2.49 0.08, 2.77 0.06, 2.89 0.13, and 3.02 0.11 (log10 of pg/ L), respectively. These mean values were signicantly higher in patients with stage II and III cancer than in healthy females (P .01 and P .0001, respectively). A trend of elevation in stage IV cancer was observed but not signicant. An ALU247-qPCR value represents the absolute amount of longer fragments of serum DNA, supposedly released from nonapoptotic dead cells. The mean logarithmically transformed ALU247-qPCR values in healthy females and patients with stage 0, I, II, III, and IV breast cancer were 1.53 0.09, 2.05 0.20, 1.54 0.11, 2.01 0.08, 2.30 0.15, and 2.55 0.12 (log10 of pg/ L), respectively. These mean values were signicantly higher in patients with stage II, III, and IV cancer than in healthy females (P .002, P .0001, P .014, respectively). The serum DNA integrity, which represents the ratio of longer DNA fragments to total serum DNA, was calculated as (ALU247qPCR value)/(ALU115-qPCR value) of each sample. The mean serum DNA integrity in healthy females and patients with stage 0, I, II, III, and IV breast cancer was 0.13 0.01 (SEM), 0.16 0.02, 0.12 0.01, 0.21 0.02, 0.29 0.03, and 0.35 0.04, respectively. The mean values were signicantly higher in patients with stage II, III, and IV breast cancer than in healthy females (P .005, P .0001, and P .002, respectively), with a clear discriminative difference (Fig 2A). The ALU115- and ALU247-qPCR values showed a trend that increased with AJCC stage. However, serum DNA integrity also
4272

showed a more prominent increase with AJCC stage. Spearmans coefcients of AJCC stage with ALU115- and ALU247-qPCR values and serum DNA integrity were 0.22 (P .04), 0.39 (P .0003), and 0.54 (P .0001), respectively. Because serum DNA integrity had the highest correlation coefcient, it was more representative of tumor progression. The ROC curve of serum DNA integrity for discriminating patients with AJCC stage II to IV breast cancer (n 51) from healthy females had an area under the curve (AUC) value of 0.79 (95% CI, 0.70 to 0.86; Fig 2B). When the specicity was set to 80%, the sensitivity for detection was 69% and the cutoff value of serum DNA integrity was 0.17. Serum DNA Integrity and Pathology Characteristics of Primary Breast Cancers Mean serum DNA integrity was independent of age in healthy females (P .18, univariate regression model) and breast cancer patients (P .20, multiple regression model with AJCC stage). Age was not a confounding factor. In the 75 patients with invasive primary breast cancer, mean serum DNA integrity and primary tumor size were signicantly correlated (r 0.48; P .0001; Fig 3). In the 82 patients with primary cancers whose lymphovascular invasion (LVI) was determined by histopathology, the mean serum DNA integrity in 34 LVI-positive tumors and 48 LVI-negative tumors was 0.25 0.02 (SEM) and 0.17 0.02, respectively. The LVI of one primary tumor was not available. Patients with LVI-positive cancers had signicantly higher mean serum DNA integrity than patients with LVI-negative cancers (P .0001); Fig 4). In 83 patients with primary breast cancer, the mean serum DNA integrity in 42 LN metastasis-positive and 41 LN metastasis-negative cancers were 0.27 0.02 and 0.14 0.02,
JOURNAL OF CLINICAL ONCOLOGY

Downloaded from jco.ascopubs.org on November 24, 2011. For personal use only. No other uses without permission. Copyright 2006 American Society of Clinical Oncology. All rights reserved.

Serum DNA Integrity in Breast Cancer Patients

Table 1. Clinicopathologic Characteristics of Primary Breast Cancers (N 83) Patients Variable Sex Male Female Tumor diameter, cm Mean SD AJCC primary tumor Tis T1 T2 T3 T4 AJCC regional lymph nodes N0 N1 N2 N3 AJCC distant metastasis M0 M1 AJCC stage 0 I II III IV Histopathologic grade G1 G2 G3 Histopathologic type DCIS Invasive ductal Invasive lobular Other Lymphovascular invasion Positive Negative No. %

0 83 3 8 32 28 10 5 41 28 10 4 80 3 8 24 27 21 3 23 26 22 8 57 13 5 34 48 2.1

0 100

10 39 34 12 6 49 34 12 5 96 4 10 29 33 25 4 32 37 31 10 69 16 6 41 58 Fig 2. (A) DNA integrity in serum from healthy females (N 51) and patients with preoperative American Joint Committee on Cancer (AJCC) stage 0 to IV breast cancer (n 8, 24, 27, 21, and 3, respectively). Bars on the right side indicate mean SEM values. Mean serum DNA integrity increased with AJCC stage (Spearmans p 0.54; P .0001). Serum DNA integrity was signicantly higher in patients with preoperative AJCC stage II to IV than in healthy females. (B) Receiver operating characteristic curve for discrimination of patients with AJCC stage II to IV breast cancer from healthy females by serum DNA integrity. Area under the curve is 0.79 (95% CI, 0.70 to 0.86).

Abbreviations: SD, standard deviation; AJCC, American Joint Committee on Cancer; DCIS, ductal carcinoma in situ. Lymphovascular invasion in one primary tumor was indenite.

respectively. Patients with LN metastasis had signicantly higher mean serum DNA integrity than patients without LN metastasis (P .0001; Fig 5A). ROC curve analysis of serum DNA integrity for discriminating patients with regional LN metastasis from patients without LN metastasis had an AUC of 0.81 (95% CI, 0.72 to 0.89; Fig 5B). When the specicity was set to 80%, sensitivity for LN metastasis prediction was 74%, and the cutoff value of serum DNA integrity was 0.18. Multivariate logistic analysis for LN metastasis status with stepwise variable selection from age, tumor size, histologic grade, LVI, estrogen receptor, progesterone receptor, HER2 level, and serum DNA integrity was performed. LVI (P .0001) and serum DNA integrity (P .0002) were signicant for LN metastasis. Serum DNA integrity was the only signicant preoperative predictor of LN metastasis.
www.jco.org

The 10 patients with micrometastasis (pN1mi) had a mean serum DNA integrity of 0.25 0.02 (SEM), signicantly higher than that of LN-negative patients (P .0001). This difference suggests that serum DNA integrity may have a clinical utility as a serum biomarker for nodal micrometastasis.
DISCUSSION

Studies have demonstrated elevated levels of free circulating DNA of various forms in serum or plasma of patients with various types of cancers.22-25 Circulating DNA as a biomarker for malignancies can be
4273

Downloaded from jco.ascopubs.org on November 24, 2011. For personal use only. No other uses without permission. Copyright 2006 American Society of Clinical Oncology. All rights reserved.

Umetani et al

Fig 3. Mean serum DNA integrity and tumor size of 75 invasive primary cancers were signicantly correlated (r 0.48; P .0001). Noninvasive tumors (eight cases of ductal carcinoma in situ) were not included in this analysis.

detected in the form of allelic imbalance, gene methylation, and gene mutation.26 Recently, it has been demonstrated that the length of cell-free plasma DNA in patients with pancreatic cancer was longer than in healthy controls.14 Another study has indicated that the integrity of circulating DNA, calculated from 400bp and 100bp qPCR threshold values of a specic gene, may be a molecular biomarker for detection of gynecologic and breast cancers.13 It is assumed that the

Fig 5. (A) Serum DNA integrity and lymph node (LN) metastasis in 42 patients with LN metastasis and 41 patients without LN metastasis. (B) Receiver operating characteristic curve for discrimination of patients with LN metastasis by serum DNA integrity. Area under the curve is 0.81 (95% CI, 0.72 to 0.89).

Fig 4. Serum DNA integrity and lymphovascular invasion status in 82 patients with assessable primary breast cancers. 4274

DNA released from malignant tumors into the bloodstream is enhanced by LVI because direct lymphatic or blood ow through the tumors enables dissemination of viable tumor cells and enhances diffusion of DNA released from dead tumor cells into the bloodstream. As a result, circulating DNA may be directly related to tumor progression and rate of tumor cell turnover, representing biologic tumor aggressiveness. Therefore, we hypothesized that integrity of circulating DNA may have a signicant prognostic utility for detection of breast cancer progression and LN metastasis. To measure integrity of circulating DNA in serum with high sensitivity and reproducibility, we developed a qPCR for ALU repeats.
JOURNAL OF CLINICAL ONCOLOGY

Downloaded from jco.ascopubs.org on November 24, 2011. For personal use only. No other uses without permission. Copyright 2006 American Society of Clinical Oncology. All rights reserved.

Serum DNA Integrity in Breast Cancer Patients

Because the ALU-qPCR assay uses serum directly as the template, it eliminates artifacts associated with DNA purication. Direct qPCR with ALU115 and ALU247 primer sets detected as little as 0.01 pg of DNA (equivalent to about 1/300 of the genome in a single cell) with high linearity.27 Such high sensitivity minimizes the required volume of template and thus makes the inhibition caused by substances in serum to become negligible. The inhibitory effect is further decreased by calculation of DNA integrity as the ratio of two almost identical qPCR assays using the same template sera and reagents (except for primers). During serum separation, cell lysis due to processing may release long DNA fragments into the serum and raise the DNA integrity. Therefore, we used a lter to remove any remaining contaminant cells from serum. Serum DNA level reportedly rises with spontaneous cell lysis if blood is not processed within 24 hours,28,29 but storage 8 hours after blood collection at room temperature did not cause significant increase of serum DNA levels with our separation protocol (data not shown). Therefore, we processed blood within 6 hours after collection. Our preliminary evaluation of 24 serum samples from patients with various neoplasms in comparison with their paired plasma indicated that only 8.2% of total DNA in serum was extraneous with our protocol; the concentration of DNA was 6.1 3.5 (mean SD) -fold higher in serum than in paired plasma after subtraction of extraneous DNA.18 In addition, using serum as a template for direct ALUqPCR was more stable and reproducible than using plasma (data not shown). Thus, serum was used for DNA integrity assessment. The absolute level of serum DNA has been demonstrated to be a potential indicator for cancer existence.7,8 Our results in this study showed that the absolute level of serum DNA measured as ALU115qPCR value was elevated, on average, in AJCC stage II and III patients. However, we found that the serum DNA integrity had a higher correlation coefcient value with tumor progression than the absolute level of serum DNA. Therefore, serum DNA integrity was considered a better representative for cancer progression. The DNA clearance rate of the patients could directly contribute to the absolute serum DNA level. In contrast, it would not signicantly
REFERENCES
1. Jemal A, Siegal R, Ward E, et al: Cancer statistics, 2006. CA Cancer J Clin 56:106-130, 2006 2. Fisher B, Bauer M, Wickerham DL, et al: Relation of number of positive axillary nodes to the prognosis of patients with primary breast cancer: An NSABP update. Cancer 52:1551-1557, 1983 3. Sanchez-Cespedes M, Esteller M, Wu L, et al: Gene promoter hypermethylation in tumors and serum of head and neck cancer patients. Cancer Res 60:892-895, 2000 4. Taback B, Hoon DS: Circulating nucleic acids and proteomics of plasma/serum: Clinical utility. Ann N Y Acad Sci 1022:1-8, 2004 5. Taback B, Hoon DS: Circulating nucleic acids in plasma and serum: Past, present and future. Curr Opin Mol Ther 6:273-278, 2004 6. Fujimoto A, ODay SJ, Taback B, et al: Allelic imbalance on 12q22-23 in serum circulating DNA of melanoma patients predicts disease outcome. Cancer Res 64:4085-4088, 2004 7. Gal S, Fidler C, Lo YM, et al: Quantitation of circulating DNA in the serum of breast cancer patients by real-time PCR. Br J Cancer 90:1211-1215, 2004 www.jco.org

inuence the values of DNA integrity because both the amounts of longer and shorter DNA fragments would be similarly affected. Unlike absolute DNA levels, which do not reect how the DNA is released, DNA integrity specically represents the relative amount of nonapoptotic cell death. Mean serum DNA integrity was signicantly higher in patients with LVI-positive tumors and had a highly signicant predictive value for LN metastasis. Even in patients with micrometastatic cancers, mean DNA integrity showed higher values than healthy females, indicating elevation with early stage of metastasis. In contrast, mean serum DNA integrity of patients with ductal carcinoma in situ was not elevated. These ndings suggest that serum DNA integrity may be helpful for preoperative evaluation of breast cancer patients with potential regional disease spreading. The prognostic value of serum DNA integrity can only be determined after long-term follow-up in earlystage patients. A multicenter study will verify the clinical utility of this assay and may reveal superiority to other established tumor markers. To evaluate the utility of serum DNA integrity as a surveillance biomarker for postoperative recurrence, we have additionally assessed 15 females with postoperative recurrence of breast cancer. The mean serum DNA integrity was 0.30 0.03 (SEM), equivalent to that of patients with stage III or IV primary breast cancer, and higher than the corresponding mean value in healthy females. This preliminary nding indicated possible utility of serum DNA integrity for surveillance of postoperative recurrence. In summary, mean serum DNA integrity was higher in sera drawn from patients with advancing stages of breast cancer than in sera drawn from healthy females. Evaluation of serum DNA integrity achieved 69% sensitivity for detection of AJCC stage II to IV primary breast cancer with a specicity of 80%, and sensitivity of 74% for LN metastasis of primary breast cancer with a specicity of 80%. This rapid and minimally invasive approach to index tumor-related circulating DNA has potential as a screening tool and for the preoperative prediction of LN metastasis.
15. Wyllie AH: Glucocorticoid-induced thymocyte apoptosis is associated with endogenous endonuclease activation. Nature 284:555-556, 1980 16. Jin Z, El-Deiry WS: Overview of cell death signaling pathways. Cancer Biol Ther 4:139-163, 2005 17. Jahr S, Hentze H, Englisch S, et al: DNA fragments in the blood plasma of cancer patients: Quantitations and evidence for their origin from apoptotic and necrotic cells. Cancer Res 61:16591665, 2001 18. Umetani N, Hiramatsu S, Hoon DSB: Higher amount of free circulating DNA in serum than in plasma is not mainly caused by contaminated extraneous DNA during separation. Annals NY Acad Sci in press, 2006 19. Hwu HR, Roberts JW, Davidson EH, et al: Insertion and/or deletion of many repeated DNA sequences in human and higher ape evolution. Proc Natl Acad Sci U S A 83:3875-3879, 1986 20. Gu Z, Wang H, Nekrutenko A, et al: Densities, length proportions, and other distributional features of repetitive sequences in the human genome estimated from 430 megabases of genomic sequence. Gene 259:81-88, 2000 21. Lander ES, Linton LM, Birren B, et al: Initial sequencing and analysis of the human genome. Nature 409:860-921, 2001 4275

8. Silva JM, Silva J, Sanchez A, et al: Tumor DNA in plasma at diagnosis of breast cancer patients is a valuable predictor of disease-free survival. Clin Cancer Res 8:3761-3766, 2002 9. Fiegl H, Millinger S, Mueller-Holzner E, et al: Circulating tumor-specic DNA: A marker for monitoring efcacy of adjuvant therapy in cancer patients. Cancer Res 65:1141-1145, 2005 10. Dulaimi E, Hillinck J, Ibanez de Caceres I, et al: Tumor suppressor gene promoter hypermethylation in serum of breast cancer patients. Clin Cancer Res 10:6189-6193, 2004 11. Muller HM, Widschwendter A, Fiegl H, et al: DNA methylation in serum of breast cancer patients: An independent prognostic marker. Cancer Res 63:7641-7645, 2003 12. Mori T, ODay SJ, Umetani N, et al: Predictive utility of circulating methylated DNA in serum of melanoma patients receiving biochemotherapy. J Clin Oncol 23:9351-9358, 2005 13. Wang BG, Huang HY, Chen YC, et al: Increased plasma DNA integrity in cancer patients. Cancer Res 63:3966-3968, 2003 14. Giacona MB, Ruben GC, Iczkowski KA, et al: Cell-free DNA in human blood plasma: Length measurements in patients with pancreatic cancer and healthy controls. Pancreas 17:89-97, 1998

Downloaded from jco.ascopubs.org on November 24, 2011. For personal use only. No other uses without permission. Copyright 2006 American Society of Clinical Oncology. All rights reserved.

Umetani et al

22. Leon SA, Shapiro B, Sklaroff DM, et al: Free DNA in the serum of cancer patients and the effect of therapy. Cancer Res 37:646-650, 1977 23. Anker P, Mulcahy H, Chen XQ, et al: Detection of circulating tumour DNA in the blood (plasma/ serum) of cancer patients. Cancer Metastasis Rev 18:65-73, 1999 24. Holdenrieder S, Stieber P, von Pawel J, et al: Circulating nucleosomes predict the response to chemotherapy in patients with advanced non-small

cell lung cancer. Clin Cancer Res 10:5981-5987, 2004 25. Herrera LJ, Raja S, Gooding WE, et al: Quantitative analysis of circulating plasma DNA as a tumor marker in thoracic malignancies. Clin Chem 51:113-118, 2005 26. Wagner PD, Verma M, Srivastava S: Challenges for biomarkers in cancer detection. Ann N Y Acad Sci 1022:9-16, 2004 27. Umetani N, Kima J, Hiramatsua S, et al: Elevated integrity of free circulating DNA in sera of

colorectal and periampullary cancer patients: A direct quantitative PCR for ALU repeats. Clinical Chemistry 52:1062-1069, 2006 28. Taback B, ODay SJ, Hoon DS: Quantication of circulating DNA in the plasma and serum of cancer patients. Ann N Y Acad Sci 1022:17-24, 2004 29. Jung M, Klotzek S, Lewandowski M, et al: Changes in concentration of DNA in serum and plasma during storage of blood samples. Clin Chem 49:1028-1029, 2003

Authors Disclosures of Potential Conicts of Interest

The authors indicated no potential conicts of interest.

Author Contributions
Conception and design: Naoyuki Umetani, Armando E. Giuliano, Dave S.B. Hoon Financial support: Dave S.B. Hoon Administrative support: Armando E. Giuliano, Dave S.B. Hoon Provision of study materials or patients: Armando E. Giuliano, Farin Amersi, Silvana Martino, Dave S.B. Hoon Collection and assembly of data: Naoyuki Umetani, Armando E. Giuliano, Suzanne H. Hiramatsu, Dave S.B. Hoon Data analysis and interpretation: Naoyuki Umetani, Suzanne H. Hiramatsu, Taku Nakagawa, Armando E. Giuliano, Dave S.B. Hoon Manuscript writing: Naoyuki Umetani, Dave S.B. Hoon Final approval of manuscript: Armando E. Giuliano, Silvana Martino, Dave S.B. Hoon

GLOSSARY

Circulating DNA: Cell-free DNA that is present in blood.

Can be detected in free form in sera or plasma. Circulating tumor DNA in the blood of cancer patients can be identical to that of the patients tumors. Therefore, circulating DNA has the potential of being a useful surrogate assay for cancer detection.

and are useful as markers for organism classication. An ALU repeat sequence is an example of a SINE.

ALU: ALU sequences are repeat sequences approximately 300

base pairs long that are found on all human chromosomes. It is a short stretch of DNA characterized by the action of the ALU restriction endonuclease. ALU insertions have been linked to many inherited diseases, including cancer.

qPCR (quantitative polymerase chain reaction): qPCR, also known as real-time PCR, consists of detecting PCR products as they accumulate. It can be applied to gene expression quantication by reverse transcription of RNA into cDNA, thus receiving the name quantitative reverse-transcriptase PCR (qRT-PCR). Despite being called quantitative, results are usually normalized to an endogenous reference. Current devices allow the simultaneous assessment of many RNA sequences. LVI (lymphovascular invasion): The presence of lymphatic or
vascular vessels invasion of a tumor. Presence of LVI in primary breast tumor increases the risk of metastasis to the regional lymph node and systemic organ sites and is therefore considered a poor prognostic factor.

SINEs (short interspersed elements): SINEs are sequences that are repeated, unblocked, and dispersed throughout the genome. They make up roughly 20% of the human genome

4276

JOURNAL OF CLINICAL ONCOLOGY

Downloaded from jco.ascopubs.org on November 24, 2011. For personal use only. No other uses without permission. Copyright 2006 American Society of Clinical Oncology. All rights reserved.