This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Carey, L. A. Articles by Sukumar, S. Search for Related Content PubMed PubMed Citation Articles by Carey, L. A. Articles by Sukumar, S. Social Bookmarking                            What's this?

Telomerase Activity and Prognosis in Primary Breast Cancers

From the Departments of Oncology, Pathology, and Surgery, The Johns Hopkins University School of Medicine, Baltimore, MD; Department of Medicine, Division of Hematology/Oncology, University of North Carolina School of Medicine, Chapel Hill, and Department of Surgery, Duke University School of Medicine, Durham, NC; and Geron Corporation, Menlo Park, CA.

Address reprint requests to Saraswati Sukumar, PhD, Breast Cancer Program, Johns Hopkins Oncology Center, Ross 370, 720 Rutland Ave, Baltimore, MD 21205-2196; email saras{at}welchlink.welch.jhu.edu

	ABSTRACT

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
PURPOSE: Recent studies associate telomerase activity with prognostic factors and survival. We compared quantitative telomerase activity in primary tumors with traditional prognostic factors and outcome in a group of invasive but nonmetastatic breast cancers.

PATIENTS AND METHODS: Telomerase activity was measured in 203 invasive breast cancers by the quantitative telomeric repeat amplification protocol method. Telomerase expression was compared with 28S rRNA level, tumor content, and clinical variables, including outcome. For clinical correlations, telomerase activity was standardized by two methods: (1) a correction for cellularity using 28S rRNA levels, and (2) a correction for the histologically determined invasive proportion of the specimen.

RESULTS: Telomerase activity was found in 82% of breast cancers with measurable 28S rRNA levels. Telomerase activity was associated with the proliferative index (P < .01) of the tumor but not with any other prognostic variable. Neither uncorrected nor corrected telomerase activity was associated with relapse-free or overall survival in this study.

CONCLUSION: Telomerase activity level was associated with the proliferative index of invasive breast cancers, but its measurement in samples from this group of nonmetastatic breast cancer patients did not predict survival.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
TELOMERASE IS A ribonucleoprotein enzyme that adds noncoding repeat sequences to the ends of linear chromosomes called telomeres. By replacing telomere ends, this enzyme counteracts normal replicative telomere loss. Active telomerase is associated with cellular immortality, including cancer cell immortality. Telomerase is active in germline cells, immortalized cell lines, and the majority of malignancies, including breast cancer.^1-4 It is generally inactive in normal somatic cells. Telomerase activity has been associated with poor survival in neuroblastoma⁵ and gastric cancer,⁶ but the relationship between the level of activity and outcome has not been widely studied. Efforts to define this relationship have been aided by recent modifications to the telomeric repeat amplification protocol (TRAP), the most commonly used assay of telomerase activity, which allow improved quantitation of telomerase activity.⁷ Using such a quantitative TRAP assay, Clark et al⁸ recently found a correlation between higher telomerase activity level and poorer overall survival in breast cancer patients with cancerous involvement of axillary lymph nodes.

Using the quantitative TRAP method for telomerase activity we examined 203 primary invasive nonmetastatic breast cancers for telomerase activity. Because telomerase is a ribonucleoprotein, stability of RNA in clinical samples could be an important factor in accurate telomerase activity quantitation. To quantify RNA stability, we used reverse transcriptase-polymerase chain reaction (RT-PCR) to measure levels of a fragment of 28S rRNA in the tumor samples, and we standardized telomerase activity to this measure of tissue integrity and cellularity. Another factor that may confound telomerase activity measurement is the natural heterogeneity of breast cancers, which contain varying amounts of normal stromal cells, inflammatory cells, and tumor cells. To correct for this, we also standardized the telomerase activity for the actual tumor content of the cancerous sample. We then compared breast cancer telomerase activity, standardized by RNA integrity, tumor cellularity, and tumor content, with traditional breast cancer prognostic factors and outcome.

	PATIENTS AND METHODS

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
Patient Population
Patients were identified by analysis of the Johns Hopkins University and Duke University Tumor Banks. Criteria for inclusion were (1) resection of primary invasive breast carcinoma between 1986 and 1991, (2) adequate frozen sample from the patient's primary tumor available for analysis, and (3) clinical variables and survival data, pertaining to each tumor, available from the tumor registries of each institution and the pathology reports. All patients meeting these criteria were included.

Tissue Samples
All tissue samples were obtained from clinical specimens stored at -80°C since resection. Institutional guidelines were observed for the acquisition and maintenance of such specimens.

Each tumor was cut as 10-µm cryosections and placed onto two or three slides. One section was stained with hematoxylin and eosin (H&E) and evaluated by the study pathologist. A section was judged to be inadequate if it had less than 10% invasive tumor, more than 50% ductal carcinoma in situ (DCIS), or exhibited necrosis. Inadequate tumors were recut until an adequate sample was obtained. If it could not be recut or remained inadequate, the tumor was removed from the study. The second 10-µm section underwent lysis and evaluation of telomerase activity by a modified TRAP method.⁷

Prognostic Factors
Clinical variables measured included age at diagnosis, gross tumor size (in centimeters), axillary lymph node status, and whether the patient had received adjuvant chemotherapy or hormonal therapy. Laboratory prognostic factors examined included measurement of the estrogen receptor (ER) content, DNA index, and proliferative index (PI) of the primary tumor. Age at breast cancer diagnosis and tumor size were coded as continuous variables. Lymph node status was coded both in a binary fashion, with a "positive" result if any lymph nodes microscopically contained tumor, and ordinally by number of lymph nodes (zero, one to three, four to nine, and > 10 involved lymph nodes). Chemotherapy and tamoxifen hormonal therapy were treated both separately (adjuvant chemotherapy, adjuvant tamoxifen) and grouped together (adjuvant therapy) and coded categorically. This is justified because meta-analysis of published adjuvant trials suggests that both adjuvant chemotherapy and adjuvant tamoxifen (in patients with ER-positive tumors) provide an approximate one third decrease in the risk of recurrence of breast cancer.^9,10 Adjuvant tamoxifen use was coded as "adjuvant therapy" only if the patient's tumor was known to be ER-positive. Some patients with ER-negative tumors received tamoxifen. Studies suggest that such patients are unlikely to have benefited from this therapy,¹⁰ so these patients were coded as "no adjuvant therapy." Patients who received tamoxifen but whose tumor ER status was unknown, were coded as "unknown" rather than as "adjuvant therapy." The ER in the primary tumor was measured with a standard biochemical assay by a commercial reference laboratory (Nichols Institute, San Juan Capistrano, CA) and coded dichotomously as "positive" or "negative." ER results were considered positive if at least 10 femtomoles of receptor per milligram of cytosolic protein was found. DNA indices were obtained from individual pathology reports and were grouped ordinally, with "diploid" defined as a DNA index of 0.90 to 1.10, "near-diploid aneuploid" defined as 1.10 to 1.30, and "aneuploid" as less than 0.90 or greater than 1.3. These cut points are derived from Dressler et al,¹¹ with the diploid range broadened because of the spectrum of reported values (L.G. Dressler, personal communication, July 1998). The PI was measured using two different methods: (1) For the tumors from the Johns Hopkins Tumor Bank, S-phase fraction was performed by a commercial laboratory according to standard flow cytometric methods (Nichols Institute, San Juan Capistrano, CA). (2) For the tumors from the Duke University tumor bank, Ki-67 was measured.¹² S-phase fraction was categorized as "low" or "high" based on the formal pathologic report from the reference laboratory, with tumors graded as low and intermediate grouped together as "low" PI. Ki-67 was categorized as "low" ( 7% staining) or "high" (> 7% staining), with this cutoff value based on the mean of the distribution among the tumors. The two PIs have been shown to correlate with each other.^13,14

Outcome Data
Outcome data were obtained from the Duke and Johns Hopkins Cancer Center Tumor Registries, with follow-up updated yearly according to cancer center protocol. These data were assessed as relapse-free survival (RFS), defined as the period from diagnosis to local or systemic relapse or last known follow-up if no relapse had occurred, and overall survival (OS), defined as the interval from diagnosis to death or last known follow-up.

Telomerase Activity Assay
Tissue extracts were prepared as previously described.¹⁵ Briefly, between 30 and 60µl of ice-cold lysis buffer (10 mmol/L Tris-HCl [pH7.5] 1mmol/L MgCl_2, 1 mmol/L ethylenebis[oxyethylene-nitrilo]tetraacetic acid, 0.1 mmol/L 4-(2-aminoethyl)-benzenesulphonyl fluoride hydrochlorine, 5 mmol/L ß-mercaptoethanol, 0.5% 3-[3-cholamidopropyl-dimethylammonio]-1-propane-sulfonate, and 10% glycerol) was added to the whole 10-µm tissue section removed from the slide by scraping with a needle. This sample was homogenized with a motorized pestle, incubated on ice for 30 minutes, then pelleted by microcentrifuging at 10,000 x g for 30 minutes at 4°C. The supernatant was removed, and the protein concentration was measured using a BCA protein determination kit (Pierce, Rockford, IL). The extracts were diluted with ice-cold lysis buffer to a final concentration of 1 mg/mL.

The quantitative TRAP assay for telomerase activity (TA) was performed as previously described,⁸ except that 0.5µg of protein (0.5- to 1-µL tissue extract) rather than 2µg of protein was assayed because histopathologically confirmed tissue sections were used rather than whole tissue (Fig 1). Quantitative TA level is expressed in telomerase product generated (TPG), defined as 1 TPG unit equals 600 molecules of TS primer¹⁵ extended at least three repeats by telomerase present in the tissue extract.

View larger version (122K): [in this window] [in a new window]   Fig 1. Representative TRAP assay of three breast cancer specimens. The characteristic six-base pair ladder of TA extension products is seen in all three samples and is absent in the heat-inactivated lanes, confirming TA enzyme activity in the samples. The 36-base pair internal control sequence used for standardization of polymerase chain reaction efficiency is seen as the lowest band in each lane. Also shown is the R8 quantitation standard and the buffer-only negative control assays, which were run with each set of breast cancer specimens.

TA Standardization
The usual method of TRAP assay standardization is by the addition of lysate that contains a prescribed amount of protein; unfortunately this method fails to correct for varying cellularity, tissue degradation, and tumor/stroma ratio. To correct for such heterogeneity, two methods were used. The first method is based on the premise that RNA degradation would coincide with loss of cell viability. The assay (RNA-corrected TA) was standardized to the amount of intact RNA present, as estimated by amplification of the 28S rRNA content (see below) of the lysate. RNA-corrected TA was defined as TA divided by the 28S rRNA levels. This provides a telomerase level standardized to the viable fraction of the sample. In the second method (histologically corrected TA), histopathologic evaluation of the H&E-stained tissue sample was performed to determine the proportion of tumor tissue within the total tissue sample. The total TA was then divided by this tumor-cell fraction to estimate the TPG per tumor tissue. Notably, this correction includes both invasive and in situ tumor cells because both can contain TA.^2,16-18 A histologic correction including only invasive tumor cells was also performed without notable difference in findings. A final correction used both RNA and histologic corrections together to estimate the TA present standardized to the amount of viable tumor tissue. This was performed by dividing the TA by the product of the 28S rRNA level times the proportion of invasive tumor tissue.

The distribution of quantitative TA can be skewed,⁸ so logarithmic transformation was performed as previously described⁸ by examining the log (corrected or uncorrected TA + 1).

28S rRNA RT-PCR
To standardize TPG results to sample cellularity, 28S rRNA levels were measured using primers designed to amplify positions 322 to 413 (EZ RNA PCR kit; Perkin-Elmer, Foster, CA), as previously described.⁸ 28S rRNA levels are expressed in cell-equivalents. The TA level (TPG) was divided by this number to obtain a quantitative TA per viable cell in the extract.

Data Analysis
The statistical analysis was performed using the JMP 3.1.6 statistical program (SAS Institute, Cary, NC). TA was measured on a continuous scale for uncorrected TA and corrected TA (RNA-corrected TA, histologically corrected TA, and both corrections together). The association between TA and continuous variables was evaluated using simple linear regression with significance testing of the regression coefficient. The association between TA and the variables measured categorically was evaluated using Wilcoxon rank sum tests.

There are no clear cut points for TA. Therefore, in outcome assessment, this variable was treated continuously and Cox's proportional hazards modeling was used to assess the contribution of the clinical variables and TA to outcome. The variables included in the multivariate Cox proportional hazards model were commonly accepted prognostic variables that were found significantly associated with outcome in univariate modeling. Of note, TA was kept in the models regardless of significance.

RFS and OS are presented as the adjusted relative risks of relapse or death associated with each prognostic variable. The relative risk associated with the continuous variables (age at diagnosis, tumor size, and TA) should be interpreted as the relative risk with each incremental unit increase in the variable. The relative risk associated with nominal variables (lymph node status, ER status, and adjuvant therapy) should be interpreted as the risk of possessing the variable compared with not possessing the variable.

	RESULTS

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
A total of 338 breast cancers were evaluated for entry onto the study. Of these, 52 were removed from the study because the sample was insufficient or contained less than 10% tumor cells in the histologically confirmed section (33 samples) or because the tumor contained more than 50% DCIS (19 samples). Thirteen samples were removed from the study for excessive thawing of the frozen tissue because this could affect telomerase enzymatic activity. Fourteen breast cancers were excluded because they came from patients with synchronous or metachronous bilateral breast cancers, which could confound outcome evaluation. An additional 21 were excluded for being from patients found to have systemic metastases at diagnosis. These exclusions left 238 patients with breast cancers (68 from Johns Hopkins and 170 from Duke University), with resections from 1986 to 1991, who had breast cancer samples frozen for research purposes and were included in the clinical databases maintained at each institution. Molecular examination of these tumors revealed that 35 lacked detectable 28S rRNA. None of the samples lacking rRNA had measurable TA, which suggests that sample degradation might be the cause of the absence of TA in these samples. These samples were excluded, leaving 203 samples for analysis. The general characteristics of these excluded tumors were not significantly different from the tumors that remained in the analysis.

General Characteristics
Patient age at diagnosis ranged from 27 to 87 years; median age was 56 years. Information regarding traditional prognostic factors is given in Table 1. Fifty-three percent of tumors were found to be ER-positive, which is at the low end of the reported range of 50% to 85%.¹⁹ Of 194 with treatment information, 103 patients (53%) had received adjuvant chemotherapy, and 63 (77%) of the 82 patients known to be ER-positive had received adjuvant tamoxifen. An additional 58 patients received tamoxifen despite unknown (34 of 48 patients) or known (24 of 72 patients) ER-negative status.

TA
Of 203 samples, 166 (82%) were positive for TA. Uncorrected TA ranged from 0 to 147.30 TPG, with a median of 2.60 TPG and an interquartile range (IQR) of 1.00 to 11.90 TPG. The distribution of TA is given in Fig 2. Uncorrected TA was marginally associated with the 28S rRNA levels (P = .08) and was not significantly associated with the proportion of tumor cells in the sample.

View larger version (16K): [in this window] [in a new window]   Fig 2. Distribution of TA in TA-positive breast cancer samples.

RNA-corrected TA, using the 28S rRNA levels, ranged from 0 to 27.35 TPG (median, 0.08 TPG; IQR 0.02 to 0.27 TPG). Histologically corrected TA, using the proportion of tumor cells in the sample, ranged from 0 to 210.43 TPG (median, 4.21 TPG; IQR 1.11 to 21.00 TPG). TA that incorporated both corrections ranged from 0 to 30.39 TPG (median 0.13 TPG; IQR 0.02 to 0.45 TPG).

Because the distribution of uncorrected and corrected TA was skewed to the right and not gaussian, all subsequent analyses used log-transformed data. In addition, the histologic correction for invasive tumor cells only did not provide any difference in results (data not shown), so all analyses presented will include the original histologic correction using the proportion of any tumor cell in the sample.

Association With Other Prognostic Factors
Examination of the relationship between TA and other prognostic factors revealed that log-transformed TA was significantly and directly associated with the PI when measured as uncorrected TA (P < .01), RNA-corrected TA (P = .01), histologically corrected TA (P = .01), and both RNA- and histologically corrected TA (P = .02). Neither uncorrected nor corrected TA was associated with the patient's age at diagnosis, size of the primary tumor, lymph node status, ER status, or if the tumor was aneuploid.

Outcome
Median follow-up was 5.5 years (range, 0.1 to 10.4 years). Eighty-four patients (41%) relapsed, and 84 patients (41%) died of any cause over the duration of the study.

Univariate Cox proportional hazard modeling revealed that positive lymph nodes and higher PI were both significantly associated with poorer RFS and OS (Table 2). The number of involved lymph nodes was also significantly associated with poorer outcome (P < .01) Larger tumor size and older age at diagnosis were associated with decreased OS. Unlike the lymph node–positive patients in the study by Clark et al,⁸ ER status did not predict either RFS or OS. Also, whether the patient had received adjuvant chemotherapy or tamoxifen did not predict either RFS or OS. The latter finding was not surprising, because it was likely that only poor prognosis patients received adjuvant therapy.

As listed in Table 2, neither uncorrected nor corrected (RNA-corrected TA, histologically corrected TA, or both corrections together) TA was associated with RFS or OS in univariate modeling. The presence of TA coded as a dichotomous variable (1 = present, 0 = absent) was also not associated with RFS (relative risk [RR] = 0.96; 95% confidence interval [CI], 0.53 to 1.85) or OS (RR = 0.88, 95% CI, 0.49 to 1.71). Evaluation of only those tumors with TA greater than 5 TPG (n = 73) also failed to reveal a relationship or a trend toward a relationship between TA and either RFS or OS.

The variables found to be associated with outcome in univariate modeling were incorporated into a multivariate model with telomerase (Table 3). Primarily, because PI was not performed routinely on all specimens at the time of diagnosis, complete data on all variables was available on 104 tumors, and, at the time of analysis, 40 patients had relapsed and 41 had died. Tumors not included in the multivariate analysis because of missing PI information did not differ in age at diagnosis, tumor size, or presence or number of involved lymph nodes from included tumors. Markedly shorter RFS and OS were strongly predicted by the presence and number of positive axillary lymph nodes at diagnosis. Higher PI in the primary tumor was independently associated with shorter RFS but not OS. Age at diagnosis remained independently associated with outcome after adjustment for tumor size and lymph node status. Tumor size, however, was no longer independently associated with outcome after adjustment for lymph node status.

Neither corrected nor uncorrected TA predicted RFS or OS. These models were examined using TA as a continuous variable, both including all samples and excluding the telomerase-negative samples, without notable changes in the results. TA was also examined as a dichotomous variable, with no independent association seen between the presence of activity and either RFS or OS.

	DISCUSSION

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
This study used an improved TRAP-based semiquantitative TA assay in 203 primary, nonmetastatic breast cancers and found that TA was associated with the tumor PI, but that it did not predict survival. Corrections for the impact of degradation, tumor cellularity, and tumor content on telomerase expression in primary breast tumors did not significantly alter our results.

Some,^3,8 but not all,^2,4 studies in breast cancer have found a correlation between telomerase levels and traditional prognostic factors. In this study, we found that quantitative TA was highly significantly associated with the PI of the primary tumor but no other prognostic factor. Previous studies have also found an association between TA and PI.^8,20,21 This association may be the result of a greater need for telomeric end maintenance in rapidly proliferating tumors to avoid critical shortening of the telomeres. More likely, however, this association reflects the fact that noncycling cells are telomerase quiescent.²²

A few studies have examined the relationship of tumor TA to survival,^5,6,20 with variable results. The most relevant is the study by Clark et al,⁸ in which a similar technique to ours was used to examine tumor characteristics, quantitative TA and RFS and OS, in a group of lymph node–positive breast cancer patients. Like this study, they detected a strong association between TA and PI. However, Clark et al also found an association between quantitative TA and OS. Their study was larger, including 395 patients with 135 relapses and 151 deaths. In that study, no association between TA and outcome was seen in the first tumor extract (2µg of protein derived from 5 mg of pulverized tissue), but an association was seen in univariate analysis between the log-transformed, RNA-corrected TA from the second extract (2µg of protein derived in the same manner from 10 mg of pulverized tissue) and RFS and OS. The log-transformed, RNA-corrected TA from the second extract remained independently associated with OS in multivariate analysis. Our study included 203 nonmetastatic patients, with 84 relapses and 84 deaths in the univariate analysis. The RRs of relapse and death in this study were not significantly associated with TA, and the trend for the association was in a protective direction (RR < 1.0). The 95% CIs for the RR of relapse (0.26 to 1.71) and death (0.28 to 1.83) do overlap the RRs found by Clark et al (RR of relapse, 1.42; of death, 1.54). Because of this overlap of CIs, the results of Clark et al and this study are not actually conflicting, and it is possible that the true effect lies somewhere within these confidence limits. Alternatively, there are several differences between the two studies that may contribute to a difference in the measured effect. The study by Clark et al included only patients with axillary lymph node metastases at diagnosis, a poor prognosis group likely representing a later stage in progression. A proportion of these patients may have had systemic metastases, an extremely poor prognosis group. Our study contained a large number of breast cancer patients without lymph node involvement, and we excluded any with systemic metastases at diagnosis (approximately 8% of our tumor bank). Secondly, we used frozen sections of breast tumor, whereas Clark et al used frozen pulverized tumor samples. By using the frozen sections instead of pulverized specimens, this study had the advantage of directly controlling the pathology and tumor content of the specimens that were used in the TRAP assay. For example, 33 tumors were excluded from our study for having less than 10% invasive cancer in the specimens, and 19 additional specimens were excluded for having more than 50% DCIS. Frozen sections, however, provide smaller samples for testing, and our levels of TA were 10-fold lower than in Clark et al.⁸ Notably, there was still no trend toward an association between TA, corrected or uncorrected, and survival when our analysis was limited to only those tumors with TA levels greater than 5 TPG.

There are several reasons to correct TA. First, breast cancers are heterogeneous, with varying proportions of nonneoplastic cells, necrosis, and preneoplastic and in situ malignant cells mixed with invasive malignant cells. The majority of the nonneoplastic cells are expected to have no TA. Thus, the usual method of standardizing the TRAP assay by protein content may not reflect the enzyme activity of the invasive tumor. We examined two methods of controlling for tissue integrity and sample cellularity. The first method used the 28S rRNA level as a measure of the sample's viability and cellularity and is similar to the method used in the study by Clark et al.⁸ By dividing the TA by the amount of RNA present in the sample, we obtained a TA corrected for tissue degradation and variability in sample cellularity. This method, however, does not distinguish between malignant and nonmalignant cells. The second method used a pathologist's determination of the proportion of the sample that was viable tumor. Division of TA by the tumor-cell proportion addressed the problem of necrotic or nonneoplastic cells in the sample. It made no difference if the proportion included invasive tumor or both invasive and noninvasive tumor. Neither of these corrections altered our results. Thus it is unlikely that tissue degradation, variable tumor cellularity, and admixed nonmalignant cells are masking a real relationship of tumor cell TA to survival. This is consistent with a recent finding from our laboratory in which TA levels from premalignant DCIS lesions were similar to TA levels from nearby invasive cancer cells.¹⁷

A finding that TA as measured by the TRAP assay does not correlate with outcome cannot be interpreted as meaning that TA is not important in vivo. The TRAP assay is a highly artificial system for evaluating telomerase function and telomere maintenance. In vivo, telomeric end maintenance is likely the result not only of telomerase enzymatic activity, but also other factors, such as telomere-associated proteins, some of which seem to have telomerase inhibitory functions.²³ If telomerase function in cancer cells does provide a selective advantage, a different model or system may be required to demonstrate it. Alternative methods may be needed for assessing telomerase expression in vivo, for example, by examining the telomerase RNA or protein component levels in the tissue sections by in situ hybridization or immunohistochemistry.

Prognostic factors in breast cancer are used in clinical decision-making regarding extent of surgery and adjuvant therapy. Unfortunately, none of the currently available markers is very accurate in prediction of outcome. Because of this uncertainty, women with breast cancer are occasionally undertreated and often overtreated. Additional markers that may aid in prediction of relapse and survival are needed. Although this study of samples from nonmetastatic breast cancer patients as described did not show correlation between TA and clinical outcome, further study of the distinction between early and late stage cancers is warranted. Future studies should examine different methods of examining the relationship between telomerase function, telomere end maintenance, and outcome.

	ACKNOWLEDGMENTS

 
Supported by United States Public Health Service grants no. CA48943 and P21CA/ES 66204, Department of Defense grant no. DAMD17-96-1-6236 (S.S.), a postdoctoral fellowship from the Susan G. Komen Foundation (C.U.), and a postdoctoral fellowship (L.A.C.) through the National Institutes of Health training grant no. CA01709 to John Fetting, MD, Johns Hopkins Oncology Center.

We thank Keith Lai, BS, and Nancy Glover, BA, for assistance with the cutting of frozen sections and data entry into the clinical database. We also express our appreciation to Ann Kammer and the staffs of the Johns Hopkins and Duke University Tumor Registry for their assistance with the clinical information and outcome data and Yuriy Gusev for his assistance with the clinical database. We thank Gary Clark for his help with the statistical analysis. Finally, we thank Dr Matthew Ewend for his helpful critiques.

	REFERENCES

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
1. Carey LA, Adey C, Henderson GH, et al: Careful histologic confirmation and microdissection reveal telomerase activity in otherwise telomerase-negative breast cancers. Clin Cancer Res4:435-440, 1998[Abstract/Free Full Text]

2. Bednarek AK, Sahin A, Brenner AJ, et al: Analysis of telomerase activity levels in breast cancer: Positive detection at the in situ breast carcinoma stage. Clin Cancer Res3:11-16, 1997[Abstract]

3. Hiyama E, Gollahon L, Kataoka T, et al: Telomerase activity in human breast tumors. J Natl Cancer Inst88:116-22, 1996[Abstract/Free Full Text]

4. Sugino T, Yoshida K, Bolodeoku J, et al: Telomerase activity in human breast cancer and benign breast lesions: Diagnostic applications in clinical specimens, including fine needle aspirates. Int J Cancer69:301-306, 1996[Medline]

5. Hiyama E, Hiyama K, Yokoyama T, et al: Correlating telomerase activity levels with human neuroblastoma outcomes. Nat Med1:249-255, 1995[Medline]

6. Hiyama E, Yokoyama T, Tatsumoto N, et al: Telomerase activity in gastric cancer. Cancer Res55:3258-3262, 1995[Abstract/Free Full Text]

7. Kim NW, Wu F: Advances in quantitation and characterization of telomerase activity by the telomeric repeat amplification protocol (TRAP). Nucleic Acids Res25:2595-2597, 1997[Abstract/Free Full Text]

8. Clark GM, Osborne CK, Levitt D, et al: Telomerase activity and survival of patients with node-positive breast cancer. J Natl Cancer Inst89:1874-1881, 1997[Abstract/Free Full Text]

9. Early Breast Cancer Trialists' Collaborative Group: Polychemotherapy for early breast cancer: An overview of the randomised trials—Early Breast Cancer Trialists' Collaborative Group. Lancet352:930-942, 1998[Medline]

10. Early Breast Cancer Trialists' Collaborative Group: Tamoxifen for early breast cancer: An overview of the randomised trials. Lancet351:1451-1467, 1998[Medline]

11. Dressler LG, Seamer LA: Standards, controls and histogram interpretation in clinical DNA flow cytometry, in Darzynkiewicz Z, Robinson P, Crissman H (eds): Methods in Cell Biology (ed 2). New York, NY, Academic Press, 1994, pp 242-262

12. Pence JC, Kizilbash AM, Kerns BJ, et al: Proliferation index in various stages of breast cancer determined by Ki-67 immunostaining. J Surg Oncol48:11-20, 1991[Medline]

13. Isola J, Kallioniemi OP, Korte JM, et al: Steroid receptors and Ki-67 reactivity in ovarian cancer and in normal ovary: Correlation with DNA flow cytometry, biochemical receptor assay, and patient survival. J Pathol162:295-301, 1990[Medline]

14. Ellis PA, Makris A, Burton SA, et al: Comparison of MIB-1 proliferation index with S-phase fraction in human breast carcinomas. Br J Cancer73:640-643, 1996[Medline]

15. Kim NW, Piatyszek MA, Prowse KR, et al: Specific association of human telomerase activity with immortal cells and cancer. Science266:2011-2015, 1994[Abstract/Free Full Text]

16. Yashima K, Milchgrub S, Gollahon LS, et al: Telomerase enzyme activity and RNA expression during the multistage pathogenesis of breast carcinoma. Clin Cancer Res4:229-234, 1998[Abstract]

17. Umbricht CB, Sherman M, Dome J, et al: Telomerase activity in DCIS associated with invasive breast cancer. Oncogene18:3407-3414, 1999[Medline]

18. Tsao J, Zhao Y, Lukas J, et al: Telomerase activity in normal and neoplastic breast. Clin Cancer Res3:627-631, 1998[Abstract]

19. Donegan WL: Tumor-related prognostic factors in breast cancer. CA Cancer J Clin47:28-51, 1997 (review) [Abstract]

20. Albanell J, Lonardo F, Rusch V, et al: High telomerase activity in primary lung cancers: Association with increased cell proliferation rates and advanced pathologic stage. J Natl Cancer Inst89:1609-1615, 1997[Abstract/Free Full Text]

21. Poremba C, Bocker W, Willenbring H, et al: Telomerase activity in human proliferative breast lesions. Int J Oncol12:641-648, 1998[Medline]

22. Holt SE, Aisner DL, Shay JW, et al: Lack of cell cycle regulation of telomerase activity in human cells. Proc Natl Acad Sci USA94:10687-10692, 1997[Abstract/Free Full Text]

23. van Steensel B, de Lange T: Control of telomere length by the human telomeric protein TRF1. Nature385:740-743, 1997[Medline]

Submitted January 15, 1999; accepted June 9, 1999.

CiteULike    Complore    Connotea    Del.icio.us    Digg    Facebook    Reddit    Technorati    Twitter    What's this?

This article has been cited by other articles:

				F. Buttitta, C. Pellegrini, A. Marchetti, A. Gadducci, S. Cosio, L. Felicioni, F. Barassi, S. Salvatore, C. Martella, G. Coggi, et al. Human Telomerase Reverse Transcriptase mRNA Expression Assessed by Real-Time Reverse Transcription Polymerase Chain Reaction Predicts Chemosensitivity in Patients With Ovarian Carcinoma J. Clin. Oncol., April 1, 2003; 21(7): 1320 - 1325. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Carey, L. A. Articles by Sukumar, S. Search for Related Content PubMed PubMed Citation Articles by Carey, L. A. Articles by Sukumar, S. Social Bookmarking                            What's this?