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 Broët, P. Articles by Spyratos, F. Search for Related Content PubMed PubMed Citation Articles by Broët, P. Articles by Spyratos, F. Social Bookmarking                            What's this?

Thymidine Kinase as a Proliferative Marker: Clinical Relevance in 1,692 Primary Breast Cancer Patients

From the Institut Curie, Paris; Assistance Publique-Hôpitaux de Marseille, Centre Hospitalier Univensitaire-Nord, Marseille; Centre P Papin, Angers; Centre R Gauducheau, St Herblain; Centre E Marquis, Rennes; Centre J Godinot, Reims; and Centre René Huguenin, St Cloud, France.

Address reprint requests to Frédérique Spyratos DrSc, Laboratoire d’Oncobiologie, Centre René Huguenin, 35 Rue Dailly, 92210 St Cloud, France; email: f.spyratos{at}stcloud-huguenin.org

	ABSTRACT

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
PURPOSE: To assess the prognostic value of thymidine kinase (TK), an enzyme involved in the DNA synthesis salvage pathway, relative to other prognostic factors in primary breast cancer.

PATIENTS AND METHODS: This retrospective study involved 1,692 patients with operable breast cancer treated in six institutions (median follow-up, 82 months). Among the 857 node-negative patients, 135 received adjuvant chemotherapy (fluorouracil, doxorubicin, cyclophosphamide [FAC] or fluorouracil, etoposide, and cisplatin [FEC]). TK was assayed in cytosol with a quantitative radioenzymatic technique. Disease-specific survival (DSS), local recurrence-free interval (LRI), and distant-relapse-free interval (DRI) were investigated.

RESULTS: High TK levels were associated with large tumor size, high histologic grade, and steroid hormone receptor negativity. Univariate analysis of the entire data set showed that high TK levels were related to shorter DSS (P < 10^-5), LRI (P < 10^-3), and DRI (P < 10^-5). In time-dependent Cox models, high TK levels remained an independent predictor of the three outcomes, both in the overall population and in node-negative patients, although its prognostic value decreased over time. In node-negative patients, the introduction of an interaction term in multivariate analysis suggested that chemotherapy was more efficacious for patients who had tumors with high TK contents. In node-positive patients, high TK levels were related only to an increased risk of LRI.

CONCLUSION: High TK values are an important risk factor in node-negative patients and seem to be associated with a beneficial effect of adjuvant FAC or FEC in patients who received adjuvant chemotherapy. The rationale of chemotherapy for patients with slowly proliferating tumors has to be discussed from a risk-benefit point of view.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
CELL KINETICS, analyzed by various methods, is an important predictor of clinical outcome in surgically treated breast cancer^1-5 and particularly in node-negative patients.^6-12 Proliferation has also been indicated as a predictor of the response to chemotherapy, including systemic adjuvant therapy^2,13 and neoadjuvant therapy,^14-16 suggesting increased chemosensitivity of rapidly proliferating tumors, even though an earlier response may not necessarily be related to longer survival.

DNA synthesis was initially measured in terms of tritiated thymidine incorporation.⁸ Cell kinetic information was subsequently obtained by static methods, such as flow cytometric DNA analysis^{1-3,5-7,9,10,12} and immunohistochemical indices, which identify a global growth fraction of cells in the G1, S, G2, and M phases of the cell cycle^4,7,9,11 or, more recently, only in the S, G2, and M phases.¹²

Another approach to measuring cell proliferation is the specific assay of enzymes involved in DNA synthesis. Thymidine kinase (TK) is an enzyme involved in the one-step salvage pathway of pyrimidine synthesis, whereas the de novo pathway synthesizes nucleotide monophosphate in multiple steps.^17-19 TK catalyzes ATP-dependent phosphorylation of thymidine to monophosphate, which is then incorporated into DNA. Its activity increases markedly after the G1-S transition and then declines rapidly in G2.¹⁹ Several TK isoenzymes are found in a given eukaryotic cell.^17,20 We focused this study on the first isoenzyme (fetal, cytosolic, TK1), which is associated with cell division, the second (adult mitochondrial TK2) being cell cycle independent. TK is conventionally assayed in terms of its enzyme activity, using radiolabeled nucleotide substrates. In human breast cancer a 14-fold increase in TK activity was found relative to normal breast tissue,²¹ and several pilot studies have shown that TK is a useful prognostic indicator in breast cancer when used in conjunction with other classic factors.^22-26

TK analysis might refine the individual prognosis and treatment of breast cancer patients. It might also form the basis for new therapeutic approaches using antimetabolites for DNA salvage pathway inhibition.²⁷

To establish the prognostic value of TK in primary breast cancer, we set up a large retrospective multicenter study involving six French institutions belonging to the Groupe d’Oncobiologie of the Fédération Nationale des Centres de Lutte Contre le Cancer. We used a radioenzymatic assay validated in conjunction with the European Organization for Research and Treatment of Cancer (EORTC) Receptors and Biomarkers Study Group.²⁸

	PATIENTS AND METHODS

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
Patients
The 1,692 patients were diagnosed and treated in the six participating centers between early 1980 and late 1994. The patients presented with unilateral nonmetastatic breast tumors and were initially treated with modified mastectomy or breast-conserving surgery.

The patients were 24 to 95 years old at diagnosis, with a mean age of 58 years. Postmenopausal status was defined as complete arrest of periods for at least 2 years; 34% of patients were premenopausal. The mean pathologic size of the tumors was 23 mm. Ductal carcinomas were diagnosed in 1,471 patients and lobular carcinomas, in 150 patients; patients with other histologic types were analyzed together. Among the 1,509 tumors graded according to Scarff et al²⁹ and Bloom and Richardson³⁰ classification, 18% were grade 1, 56% were grade 2, and 26% were grade 3. A mean of 13 lymph nodes were examined. Fifty-seven percent of patients were node-negative and 26% had one to three invaded nodes. Estrogen and progesterone receptor positivity was observed in 71% and 62% patients, respectively (Table 1). The primary treatment was tumorectomy (64%) or mastectomy (36%) with axillary dissection, followed by radiotherapy in 85% of cases. Data on adjuvant treatment were available for 1,595 patients. A total of 649 patients received no adjuvant treatment, whereas 302 received chemotherapy and 411 received hormone therapy with tamoxifen (20 mg daily) for at least 2 years. Two hundred thirty-three patients received both treatments. Among node-positive patients, 602 received adjuvant treatment; chemotherapy protocols varied widely among the centers, but 70% of patients received anthracycline-containing regimens and 90% received fluorouracil.

Tumor Material
Tumor samples were histologically examined by a pathologist at the time of initial surgery and were subsequently used for routine determination of ER and PgR by biochemical methods. For TK assay, three centers used cytosols prepared for hormone receptor assays at the time of initial surgery and kept in liquid nitrogen (centers 4 and 5) or at -80°C (center 6). The other centers prepared cytosols specifically for this study by homogenizing fragments of frozen tumors in the same hormone receptor buffer (Tris HCl 10 mmol/L, EDTA 1.5 mmol/L, Na₂MoO₄ 10 mmol/L, dithiothreitol 0.5 mmol/L, and 10% glycerol; pH 7.4). After centrifugation for 1 hour at 105,000 x g, supernatants were aliquoted and stored until analysis (maximum 6 months) in liquid nitrogen (center 2) or at -80°C (centers 3 and 1). The cytosol protein concentration was determined for this study by using the same BCA assay (Pierce Chemical, Rockford, IL) and the same standard for all of the samples.

ER and PgR Assays
The hormone receptor status of the tumor was recorded at the time of surgery. The tumor was considered to be steroid hormone receptor–positive if the ER or PgR values exceeded 10 fmol·mg^-1 protein in radio ligand assay or 15 fmol·mg^-1 protein in enzyme immune assay. Quality control was based on regular testing of both internal and EORTC controls.^31,32

TK Assay
TK enzyme activity was determined using the Prolifigen TK radioenzymatic assay(Sangtec Medical, Bromma, Sweden) as previously described,²⁸ with the following modifications recommended by the EORTC Receptors and Biomarkers Study Group. Cytosols were diluted 1/20 in heat-inactivated normal calf serum containing HEPES 90 mmol/L, potassium chloride 18 mmol/L, dithiothreitol 6 mmol/L, MgCl₂ 8 mmol/L, and ATP 4 mmol/L. TK levels were expressed in mU.·mg^-1 protein. All of the laboratories used a standardized protocol and a quality control program³³ designed by three laboratories (centers 1, 2, and 6), focusing particularly on the impact of storage conditions. Between-run coefficients of variation ranged from 12.4% to 15.9%.

Statistical Methods
TK levels were compared with clinical, biologic, and histologic features by using parametric and nonparametric tests as appropriate.³⁴ To investigate the influence of sample storage conditions on TK measurements dispersion, we tested homogeneity of variances using a one-way analysis of variance test based on absolute deviation from medians.³⁵ Two-sided P values below .05 were considered significant.

Disease-specific survival (DSS), local-relapse-free intervals (LRIs), and distant-relapse-free intervals (DRIs) were calculated from the date of first treatment. Disease-related deaths were scored as events, and other patients were censored at the time of cancer-unrelated death or last follow-up. Local relapse was defined as a tumor arising in the treated breast or regional lymph nodes. Patients who sustained a distant relapse as the first event were still considered at risk for experiencing a local relapse.

Survival curves were derived from Kaplan-Meier estimates.³⁶ The log-rank test was used to test the equality of survival distributions.³⁷ Survival rates and relative risks are presented with their 95% confidence intervals. For univariate analysis to examine the relationship between TK and outcome, we used three prespecified cut points in each participating center, corresponding to the 25th, 50th (or median), and 75th percentiles of the TK distribution.

For the three outcomes, the following variables were tested by multivariate analysis using a Cox proportional hazards regression model³⁸: age at first treatment (40 or younger, 41 to 60, and older than 60 years), menopausal status (premenopausal v postmenopausal), pathologic tumor size ( 20 mm or > 20 mm), lymph node involvement (none, one to three, or four or more), histologic grade, histologic type (ductal, lobular, or other), and steroid hormone receptor status. Because the distribution of TK values was highly positively skewed, this marker was log transformed when TK was tested as a continuous variable.

Because node-negative and node-positive patients are usually investigated separately in clinical studies, and because the effect of TK could be different in these two groups, we separately analyzed the two categories. Because the clinical management of node-positive patients was quite heterogeneous, the treatment effect was only investigated in node-negative patients.

To address the question of a potential interaction between TK levels and chemotherapy in the node-negative patient population (for which the chemotherapy regimen was homogeneous), we introduced and tested an interaction term in the final multivariate Cox models. The interaction variable corresponded to the product of the TK value by the treatment covariate (chemotherapy v no chemotherapy).

Confounding variables with k subgroups were coded to (k-1) dummy variables (this model permits non–log-linear relation of relative risks between two subsequent subgroups when k > 2). The prognostic role of TK was analyzed in a two-stage procedure. First, the clinical and biologic variables described above were entered in a stepwise forward procedure. Second, to test the additional prognostic information carried by TK, this marker was entered in the final model considered previously as a continuous variable (with logarithmic transformation). To address the problem of variations among centers, we stratified the multivariate analyses according to the center. For LRI, we also stratified the analyses according to the type of surgery (mastectomy v tumorectomy).

Moreover, to identify a possible time-varying effect of TK, we also tested a time-dependent variable for TK. In this time-dependent Cox model, the form of dependence was

equation

where was the hazard rate; ₀ was the baseline hazard; t was the time variable; ß, , and were regression coefficients; TK was the TK value; and Z was the vector of other prognostic factors. This model allowed for the possibility that the effect of TK was not constant throughout follow-up (nonproportional hazards). All analyses were performed with the S-Plus version 3.2 software package.³⁹

	RESULTS

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
The median TK value in the entire population was 69 mU·mg^-1 protein (range among centers, 42 to 140 mU·mg^-1 protein). Significant differences in mean TK values were observed according to the center and storage conditions (Table 2). Higher TK levels were observed when cytosols were stored in liquid nitrogen (centers 2, 4, and 5) than at -80°C (P < 10^-4) and when they were stored for less than 1 year (centers 1, 2, and 3) than for more than 1 year (P < 10^-4). No significant difference in the variance of TK values was observed according to the center or storage conditions.

In the entire population, the disease-related survival rate was 85.5% (± 0.9%) at 60 months and 71.1% (± 1.6%) at 120 months. The proportions of patients free of local and distant relapse were 91.6% (± 0.7%) and 80.7% (± 1.1%), respectively, at 60 months, and 85.9% (± 1.3) and 70.2% (± 1.3%) at 120 months.

Univariate and Multivariate Analyses of the Entire Data Set
Univariate analysis indicated that high TK levels were significantly linked to a worse prognosis, in terms of DSS (P < 10^-5), LRI (P < 10^-3), and DRI (P < 10^-5). Five-year rates are listed for the different percentiles in Table 4.

View larger version (13K): [in this window] [in a new window]   Fig 1. (A) DSS curves of overall population according to TK-prespecified percentiles (P < 10-5). (B) LRI survival curves of overall population according to TK-prespecified percentiles (P < 10-3). (C) DRI survival curves of overall population according to TK-prespecified percentiles (P < 10-5).

View larger version (13K): [in this window] [in a new window]   Fig 2. DRI survival curves in node-negative patients according to TK-prespecified percentiles (P < 10-6).

In multivariate analysis (Table 6), the risk of disease-related death increased with high TK values (P = 10^-3) after adjustment for tumor size (P = 10^-2) and PgR status (P = .004). The risk of local relapse increased with high TK values and age younger than 40 years. The risk of distant relapse increased with high TK values, young age, and tumor size larger than 20 mm. For the three outcomes, histologic grading was removed from the model when TK measurement was considered as a candidate prognostic factor in the stepwise analysis.

Interaction between chemotherapy and TK on DSS and DRI. Among the 857 node-negative patients, 135 (16%) received adjuvant chemotherapy alone, 524 (61%) received neither chemotherapy nor hormone therapy, 171 (20%) received hormone therapy, and 27 (3%) received a combination of hormone therapy and chemotherapy. At the end of the study, among patients who did and did not receive adjuvant chemotherapy, there were 34 (25%) and 86 (16%) cases of metastasis, respectively, and 25 (19%) and 67 (13%) deaths, respectively.

In univariate analysis, the TK level was not significantly linked to a worse DSS rate or a shorter DRI (Fig 3) among the 135 patients who received adjuvant chemotherapy relative to patients who received no adjuvant treatment. Considering the DRI, relative risk between the lowest quartile and the highest quartile was close to 1 in the treated group, whereas it was approximately 4 in the untreated group (receiving neither chemotherapy nor hormone therapy). In the treated group, 85.4% (± 7.0%) of patients with TK values below the 25th percentile were free of distant relapse at 5 years, compared with 82.3% (± 6.0%) of those with values above the 75th percentile. In the untreated group, 94.3% (± 2.0%) of patients with TK values below the 25th percentile were free of distant relapse at 5 years, compared with 72.1% (± 4.5%) of patients with values above the 75th percentile.

View larger version (13K): [in this window] [in a new window]   Fig 3. DRI survival curves in node-negative patients receiving adjuvant chemotherapy according to TK-prespecified percentiles (P = not significant).

Univariate and Multivariate Analyses of Node-Positive Patients
In univariate analysis, TK was not significantly linked to worse DSS or a shorter DRI. In contrast, high TK levels were significantly linked to a shorter LRI (P = .02). The DRIs for the different percentiles are shown in Fig 4.

View larger version (13K): [in this window] [in a new window]   Fig 4. DRI survival curves in node-positive patients according to TK-prespecified percentiles (P = not significant).

	DISCUSSION

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
TK was assayed in a series of 1,692 primary breast cancer specimens to investigate its prognostic significance relative to other factors. This multicenter study (six French institutions) obviously had inherent advantages and pitfalls. Among the advantages was the large number of patients with available clinical, biologic, and histologic information. Conversely, the population was heterogeneous, with regional differences in breast cancer management. This led to cautious interpretation of potential interactions between TK levels and adjuvant chemotherapy.

Significant differences in the population characteristics among the participating centers were observed. For example, two centers (centers 3 and 4) recruited only node-negative patients for this study. TK was assayed in laboratories that have used precise standardized methods and a quality control program for a number of years.^28,33 The variations observed from one center to another are likely not related to differences in the way the assay was performed. Indeed, as part of the quality control process, identical specimens were assayed in several centers and the results were not significantly different from one center to another. Storage variability, a crucial factor in multicenter studies, was observed in this series. The lowest mean of TK values was obtained in center 6, which used cytosols kept for several years at -80°C, and the highest mean was obtained in center 2, which specifically prepared cytosols for this study and kept them in liquid nitrogen for less than 6 months. However, the dispersion of TK values was homogeneous among centers. To take into account dissimilarities in storage conditions and recruitment, each quartile used in univariate analysis gathered all the data from each center with respect to its own considering quartile. In multivariate analyses, adjustments for the center were made at each step of the statistical procedure.

TK levels were independent of menopausal status and age but were higher in patients younger than 40 years of age, in keeping with a previous report indicating higher proliferative rates in patients younger than 40 years.⁴⁰ In our series, the difference was not statistically significant, however, probably because of the small number of patients younger than 40. High TK values were linked to large pathologic tumor size, in agreement with previous studies.^24,25 We also confirmed the negative correlation of TK with ER^23,24 and PgR status^22,24,25 and the positive correlation with histologic grade.^22,24,26

In the overall population, univariate analysis showed that higher TK values were related to shorter survival (DSS, LRI, and DRI), confirming previous reports.^22,24,25 In Cox multivariate analyses, TK was an independent prognostic factor for all three outcomes in the overall population and among node-negative patients, both in the overall node-negative population and in the node-negative patients who did not receive adjuvant chemotherapy (not shown). TK was only predictive of the LRI in node-positive patients.

It is noteworthy that TK values and histologic grade were strongly intercorrelated and yielded common prognostic information. Histologic grade has already been found to be of prognostic importance in node-negative patients.⁴¹ In our node-negative patients, TK provided additional information beyond that given by histologic grade, whereas TK did not add further information in node-positive patients.

Thus our findings confirm the potential clinical utility of proliferation parameters, as already shown using the labeling index,⁸ flow cytometry,^5-7,9,10 immunohistochemistry,^9,11,12 and TK assay.²² Compared with other methods of measuring proliferation, TK has certain advantages in that it (1) avoids the exclusion of many cases, such as S-phase fraction measurement by flow cytometry and (2) is a truly quantitative method suited to quality control programs, whereas antigen retrieval and immunostaining techniques have not yet been standardized for immunohistochemical proliferation markers.

We used a time-dependent Cox model to investigate the prognostic value of TK. Cox proportional hazards regression is used in most clinical studies of prognostic factors, but few studies based on this approach have reported verification of the so-called proportional hazards assumption. A loss of prognostic significance with longer follow-up can be due to real changes in the effect of a factor as a function of time. Such time-dependent variation would violate the proportional hazards assumption inherent in Cox regression. Failure to recognize such a violation can lead to false estimates of prognostic value,^42,43 a problem encountered in studies investigating other proliferation-related markers such as S phase.^10,42,43 Using a time-dependent Cox model, we showed that the prognostic value of TK decreased as follow-up increased. This time-varying effect of TK could be explained by the fact that TK reflects tumor growth. Thus early relapses would be more likely to occur in patients whose tumors have higher TK values than in those with lower TK levels, and the relative risk between these two groups would vary with time.

Studies based on flow cytometric DNA analysis have suggested that highly proliferative tumors have increased sensitivity to chemotherapy in neoadjuvant^14-16 and adjuvant protocols,^2,13 regardless of the impact on survival. We investigated the relation between TK and chemotherapy in node-negative patients who received homogeneous protocols (FAC or FEC). Chemotherapy was obviously given to high-risk patients (high SBR grade, negative hormone receptors, and so on). When confounding factors were taken into account in multivariate analyses, the introduction of an interaction term suggested that chemotherapy was more effective in node-negative patients whose tumors had high TK contents. This possibility has already been raised for the S phase; high rates of proliferation seem to predict the response to chemotherapy. As the interaction for chemotherapy and TK was not the primary focus of this study, these findings should be interpreted with care; indeed, they may be due to uncontrolled factors related to TK and chemotherapy.

Conversely, the rationale of chemotherapy for patients with a good prognosis and slowly proliferating tumors, as shown by TK assay in our study or by other methods, may be also reassessed in terms of the likely risk-benefit ratio. This was already pointed out by the results of the San Antonio Specialized Program of Research Excellence breast cancer database⁴⁴ and those of Bryant et al¹⁰ in patients with node-negative, ER-positive tumors. Randomized, well-controlled studies of homogeneous chemotherapy protocols are required to settle the issue.

In node-positive patients, we did not investigate the interaction between TK and adjuvant chemotherapy, due to great heterogeneity in the chemotherapy protocols; however, the impact of chemotherapy could explain the lack of prognostic significance of TK in this subpopulation.

Finally, we found that the TK level was an independent prognostic factor for local recurrence. This is consistent with reports that high-labeling index predicted early recurrence in the overall population and in axillary node-negative breast cancer patients.⁴⁵ Few predictive factors of local recurrence have been identified, and our results could be of potential clinical relevance when discussing the type of surgical treatment for patients with high TK values. However, local relapse is likely to be followed by disseminated disease, on which more radical surgery would be ineffective.

In conclusion, these findings could provide the basis for future studies of the clinical utility of TK assay in breast cancer, particularly in node-negative patients. Proliferation indexes such as TK could be used to adjust adjuvant treatment to individual tumor characteristics.

	ACKNOWLEDGMENTS

 
We thank all the pathologists and clinician teams of the different centers who actively participated in the study. We are grateful for the expert technical assistance of the persons involved in the TK assays in each center.

	REFERENCES

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
1. Stal O, Dufmats M, Hatschek T, et al: S-phase fraction is a prognostic factor in stage I breast carcinoma. J Clin Oncol 11: 1717-1722, 1993[Abstract/Free Full Text]

2. Kute TE, Quadri Y, Muss H, et al: Flow cytometry in node-positive breast cancer: Cancer and Leukemia Group B protocol 8869. Cytometry 22: 297-306, 1995[Medline]

3. Camplejohn RS, Ash CM, Gillett CE, et al: The prognostic significance of DNA flow cytometry in breast cancer: Results from 881 patients treated in a single centre. Br J Cancer 71: 140-145, 1995[Medline]

4. Thor AD, Liu S, Moore DH II, et al: Comparison of mitotic index, in vitro bromodeoxyuridine labeling, and MIB-1 assays to quantitate proliferation in breast cancer. J Clin Oncol 17: 470-477, 1999[Abstract/Free Full Text]

5. O’Reilly SM, Camplejohn RS, Barnes DM, et al: Node-negative breast cancer: Prognostic subgroups defined by tumor size and flow cytometry. J Clin Oncol 8: 2040-2046, 1990[Abstract]

6. Witzig TE, Ingle JN, Cha SS, et al: DNA ploidy and the percentage of cells in S-phase as prognostic factors for women with lymph node negative breast cancer. Cancer 74: 1752-1761, 1994[Medline]

7. Dettmar P, Harbeck N, Thomssen C, et al: Prognostic impact of proliferation-associated factors MIB1 (Ki-67) and S-phase in node-negative breast cancer. Br J Cancer 75: 1525-1533, 1997[Medline]

8. Silvestrini R, Daidone MG, Luisi A, et al: Cell proliferation in 3,800 node-negative breast cancers: Consistency over time of biological and clinical information provided by 3H-thymidine labelling index. Int J Cancer 74: 122-127, 1997[Medline]

9. Brown RW, Allred CD, Clark GM, et al: Prognostic value of Ki-67 compared to S-phase fraction in axillary node-negative breast cancer. Clin Cancer Res 2: 585-592, 1996[Abstract]

10. Bryant J, Fisher B, Gunduz N, et al: S-phase fraction combined with other patient and tumor characteristics for the prognosis of node-negative, estrogen-receptor-positive breast cancer. Breast Cancer Res Treat 51: 239-253, 1998[Medline]

11. Clahsen PC, van de Velde CJ, Duval C, et al: The utility of mitotic index, oestrogen receptor and Ki-67 measurements in the creation of novel prognostic indices for node negative breast cancer. Eur J Surg Cancer 25: 356-363, 1999

12. Rudolph P, Alm P, Heidebrecht HJ, et al: Immunologic proliferation marker Ki-S2 as prognostic indicator for lymph node-negative breast cancer. J Natl Cancer Inst 3: 271-278, 1999

13. Stal O, Skoog L, Rutqvist LE, et al: S-phase fraction and survival benefit from adjuvant chemotherapy or radiotherapy of breast cancer. Br J Cancer 70: 1258-1262, 1994[Medline]

14. Spyratos F, Briffod M, Tubiana-Hulin M, et al: Sequential cytopunctures during preoperative chemotherapy for primary breast carcinoma. II. DNA flow cytometry changes during chemotherapy, tumor regression, and short-term follow-up. Cancer 69: 470-475, 1992[Medline]

15. Remvikos Y, Mosseri V, Asselain B, et al: S-phase fractions of breast cancer predict overall and post-relapse survival. Eur J Cancer 33: 581-586, 1997

16. Chevillard S, Lebeau J, Pouillart P, et al: Biological and clinical significance of concurrent p53 gene alterations, MDR1 gene expression, and S-phase fraction analyses in breast cancer patients treated with primary chemotherapy or radiotherapy. Clin Cancer Res 3: 2471-2478, 1997[Abstract/Free Full Text]

17. Hannigan BM, Barnett YA, Armstrong DB, et al: Thymidine kinases: The enzymes and their clinical usefulness. Cancer Biother 8: 189-197, 1993[Medline]

18. Petty EM, Miller DE, Grant AL, et al: FISH localization of the soluble thymidine kinase gene (TK1) to human 17q25, a region of chromosomal loss in sporadic breast tumors. Cytogenet Cell Genet 72: 319-321, 1996[Medline]

19. Sherley JL, Kelly TJ: Regulation of human thymidine kinase during the cell cycle. J Biol Chem 15: 8350-8358, 1988

20. Soderlund JC, Arner ES: Mitochondrial versus cytosolic activities of deoxyribonucleoside salvage enzymes. Adv Exp Med Biol 370: 201-204, 1994[Medline]

21. Sakamoto S, Ebuchi M, Iwama T: Relative activities of thymidylate synthetase and thymidine kinase in human mammary tumours. Anticancer Res 13: 205-207, 1993[Medline]

22. Spyratos F, Martin PM, Hacene K, et al: Multiparametric prognostic evaluation of biological factors in primary breast cancer. J Natl Cancer Inst 84: 1266-1272, 1992[Abstract/Free Full Text]

23. O’Neill KL, Hoper M, Odling-Smee GW: Can thymidine kinase levels in breast tumors predict disease recurrence? J Natl Cancer Inst 84: 1825-1828, 1992[Abstract/Free Full Text]

24. Romain S, Christensen IJ, Chinot O, et al: Prognostic value of cytosolic thymidine kinase activity as a marker of proliferation in breast cancer. Int J Cancer 61: 7-12, 1995[Medline]

25. Romain S, Martin PM, Klijn JG, et al: DNA-synthesis enzyme activity: A biological tool useful for predicting anti-metabolic drug sensitivity in breast cancer? Int J Cancer 74: 156-161, 1997[Medline]

26. Wafflart J, Daver A, Grenier J, et al: Cytosolic thymidine kinase activity as a marker of cell proliferation in primary breast cancer. Breast Cancer Res Treat 14: 135, 1989 (abstr)

27. Vazquez-Padua MA: Modulation of thymidine kinase activity: A biochemical strategy to enhance the activation of antineoplastic drugs. P R Health Sci J 13: 19-23, 1994[Medline]

28. Romain S, Spyratos F, Guirou O, et al: Technical evaluation of thymidine kinase assay in cytosols from breast cancers: EORTC Receptor Study Group Report. Eur J Cancer 30A: 2163-2165, 1994

29. Scarff RW, Torloni H: Histological Typing of Breast Tumors, No. 13-20. Geneva Switzerland: World Health Organization, 1968

30. Bloom HJ, Richardson WW: Histological grading and prognosis in breast cancer. Br J Cancer 11: 359-377, 1957[Medline]

31. Revision of the standards for the assessment of hormone receptors in human breast cancer: Report of the second E.O.R.T.C. Workshop, 16-17 March, 1979, in the Netherlands Cancer Institute. Eur J Cancer 16:1513-1515, 1980

32. Romain S, Spyratos F, Goussard J, et al: Improvement of quality control for steroid receptor measurements: Analysis of distributions in more than 40000 primary breast cancers—French Study Group on Tissue and Molecular Biopathology. Breast Cancer Res Treat 41: 131-139, 1996[Medline]

33. Spyratos F, Romain S, Rostaing-Puissant B, et al: Standardization and quality control in the evaluation of proliferation parameters in T1T2, N0N1, M0 breast cancer: Multicentric retrospective study— I. DNA synthesis enzyme activities. Bull Cancer 86: 678-684, 1999[Medline]

34. Conover WJ: Practical Nonparametric Statistics (ed 2). New York, NY: Wiley, 1980

35. Brown MB, Forsythe AB: Robust tests for the equality of variances. J Am Stat Assoc 53: 457-481, 1974

36. Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53: 457-481, 1958

37. Peto R, Peto J: Asymptotically efficient rank invariant test procedure. J R Stat Soc A 135: 185-207, 1972

38. Cox DR: Regression models and life-tables. J R Stat Soc 34: 187-220, 1972

39. S-Plus Version 3.2: Guide to Statistics. Seattle WA: Statistical Sciences, 1993

40. Remvikos Y, Magdelenat H, Dutrillaux B: Genetic evolution of breast cancers: III. Age-dependent variations in the correlations between biological indicators of prognosis. Breast Cancer Res Treat 34: 25-33, 1999

41. Le Doussal V, Tubiana-Hulin M, Hacene K, et al: Nuclear characteristics as indicators of prognosis in node negative breast cancer patients. Breast Cancer Res Treat 14: 207-216, 1989[Medline]

42. Gamel JW, Meyer JS, Province MA: Proliferative rate by S-phase measurement may affect cure of breast carcinoma. Cancer 76: 1009-1018, 1995[Medline]

43. Hilsenbeck SG, Ravdin PM, de Moor CA, et al: Time-dependence of hazard ratios for prognostic factors in primary breast cancer. Breast Cancer Res Treat 52: 227-237, 1998[Medline]

44. Wenger CR, Clark GM: S-phase fraction and breast cancer: A decade of experience. Breast Cancer Res Treat 51: 255-265, 1998[Medline]

45. Silvestrini R, Daidone MG, Luisi A, et al: Biologic and clinicopathologic factors as indicators of specific relapse types in node-negative breast cancer. J Clin Oncol 13: 697-704, 1995[Abstract/Free Full Text]

Submitted March 15, 2000; accepted March 7, 2001.

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

This article has been cited by other articles:

				M. Colozza, E. Azambuja, F. Cardoso, C. Sotiriou, D. Larsimont, and M. J. Piccart Proliferative markers as prognostic and predictive tools in early breast cancer: where are we now? Ann. Onc., November 1, 2005; 16(11): 1723 - 1739. [Abstract] [Full Text] [PDF]

				A. Ohrvik, M. Lindh, R. Einarsson, J. Grassi, and S. Eriksson Sensitive Nonradiometric Method for Determining Thymidine Kinase 1 Activity Clin. Chem., September 1, 2004; 50(9): 1597 - 1606. [Abstract] [Full Text] [PDF]

				D L Francis, A Freeman, D Visvikis, D C Costa, S K Luthra, M Novelli, I Taylor, and P J Ell In vivo imaging of cellular proliferation in colorectal cancer using positron emission tomography Gut, November 1, 2003; 52(11): 1602 - 1606. [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 Broët, P. Articles by Spyratos, F. Search for Related Content PubMed PubMed Citation Articles by Broët, P. Articles by Spyratos, F. Social Bookmarking                            What's this?