Real-time quantitative RT PCR and detection of tumour cell dissemination in breast cancer patients: plasmid versus cell line dilutions

Transcription

1 Original article Annals of Oncology 14: , 2003 DOI: /annonc/mdg341 Real-time quantitative RT PCR and detection of tumour cell dissemination in breast cancer patients: plasmid versus cell line dilutions M. Saad Ismail 1,2, W. Wynendaele 1, J. L. E. Aerts 3, R. Paridaens 1 *, L. Van Mellaert 4, J. Anné 4, R. Gaafar 2, N. Shakankiry 2, H. M. Khaled 2, M. R. Christiaens 1, S. Omar 2, P. Vandekerckhove 3 & A. T. van Oosterom 1 1 Department of Oncology, UZ Gasthuisberg, Leuven, Belgium; 2 National Cancer Institute, Cairo, Egypt; 3 Experimental Laboratory Medicine, UZ Gasthuisberg, Leuven; 4 Rega Institute, K.U. Leuven, Leuven, Belgium Received 25 June 2002; revised 9 December 2002; accepted 4 April 2003 Background: We previously developed a real-time quantitative RT PCR technique to detect breast carcinoma cells in peripheral blood (PB). The aim of the current study was to improve cytokeratin 19 (CK19) quantification using plasmid dilutions of cloned PCR fragments to obtain a more reliable and reproducible quantification of CK19 transcripts. Materials and methods: PB samples of 14 stage IV breast cancer patients and 23 healthy controls were examined with RT PCR using plasmid quantification. Results: Median CK19+ copy numbers of one and 11 were detected in the control group and stage IV breast cancer patients, respectively (Mann Whitney, P ). When comparing the results obtained using cell line dilutions with those obtained using plasmid dilutions, a good correlation was observed (r 2 = 0.98). Conclusions: Plasmid dilutions are more reliable than cell line dilutions for quantification of gene expression, and more objective criteria for positivity could be defined based on the characteristics of the standard curve (slope and intercept). A more universally accepted agreement on the definition of the cut-off value for positivity is needed. Key words: breast cancer, cytokeratin 19, plasmid dilutions, quantitative RT PCR Introduction Several prospective studies have confirmed the clinical importance of occult tumour cells in the bone marrow of breast cancer patients, representing an independent predictive and prognostic factor for distant relapse and overall survival [1 3]. However, little is known about the natural history of micrometastases. Many groups have suggested that monitoring of minimal residual disease could be used to improve disease staging, to assess treatment response in individual patients or as a marker for evaluating new therapeutic strategies [1 10]. Peripheral blood (PB) is an attractive source of samples due to the ease of obtaining blood from patients. We previously developed a real-time quantitative RT PCR technique to detect breast carcinoma cells in PB. This technique is sensitive and has a high reproducibility with many advantages over classic quantitative PCR methods. We detected significantly elevated levels of cytokeratin 19 positive (CK19+) cells in PB of These three authors contributed equally to the development of this work. *Correspondence to: Professor R. Paridaens, Gezwelziekten, Laboratorium Experimentele Oncologie, UZ Gasthuisberg, Herestraat 49, 3000 Leuven, Belgium. Tel: ; Fax: ; robert.paridaens@uz.kuleuven.ac.be <10% of the volunteers, in ±30% of stage I III patients and in >70% of stage IV breast cancer patients. We also applied the same technique to bone marrow samples of stage I IIIa breast cancer patients [5]. In the current study, the aim was to further validate our technique and to improve CK19 quantification using plasmid dilutions of the cloned PCR fragment to obtain a more reliable and reproducible quantification of CK19 transcripts. Materials and methods Patients and samples Peripheral blood samples of 14 stage IV breast cancer patients and 23 healthy controls, as described previously, were re-analysed. Sample processing, RNA extraction and cdna synthesis We have previously described these procedures extensively. In brief, starting from RNAzol lysates, total RNA was extracted using chloroform, precipitated with isopropanol and washed with 70% ethanol. The resulting pellet was redissolved in nuclease-free water. RNA concentrations were measured using a spectrophotometer (260 nm/280 nm). After heating at 65 C for 5 min to denature RNA and to inactivate RNases, 1 µg total RNA was subjected to reverse transcription using 300 U M-MLV Reverse Transcriptase (Life Technologies, Gaithersburg, MD, USA), 30 U RNasin RNase inhibitor (Promega, 2003 European Society for Medical Oncology

2 1242 Madison, WI, USA), 25 µm random hexamer primers and RT buffer containing 250 mm Tris HCl ph 8.3, 375 mm KCl and 15 mm Mg 2+ in a total volume of 40 µl at 37 C for 2 h. The reaction was terminated by heating at 65 C for 10 min. Plasmid construction, amplification and purification A 101-bp CK19-PCR fragment generated using primers, as described by Slade et al. [11], was cloned in the pgem -T Easy vector (Promega) and introduced in Escherichia coli TG1. From a selected transformant containing the desired construct, plasmid DNA was isolated using the Wizard Plus SV Miniprep DNA purification system (Promega). The resulting CK19 plasmid was subsequently linearised by NcoI digestion. In addition, PCR fragments for two control genes, β2-microglobulin (β2m) and β-glucuronidase (GUS), were cloned in the pcr II-TOPO vector (Invitrogen, Carlsbad, CA, USA; Life Technologies). To purify both plasmids, the same technique as for CK19 was applied. The resulting β2m and GUS plasmids were linearised by BamHI and HindIII digestion, respectively. Dilutions of linearised plasmids were prepared in 1 mm Tris HCl ph 8.0, 0.1 mm EDTA containing 50 µg/ml E.coli trna. Serial 10-fold dilutions were made in the range of copies to one copy. PCR For each PCR, 6 µl cdna (diluted 1:3 in nuclease-free water) or plasmid product (serial dilutions), 25 µl Universal PCR Master Mix (Applied Biosystems, Foster City, CA), 900 nm forward primer, 900 nm reverse primer, 200 nm probe and nuclease-free water were added to a final volume of 50 µl. Amplification and detection were performed with the ABI Prism 7700 sequence detection system (Applied Biosystems). The thermal cycle used was 2 min at 50 C, 10 min at 95 C and 50 cycles of 15 s denaturation at 95 C and 1 min annealing at 60 C. Quantification Quantification was based on the Taqman principle. Before annealing to the target sequence, the fluorescence of the dual labelled (one fluorochrome, one quencher molecule) oligonucleotide probe is quenched. Upon hybridisation of the probe to the target sequence, the probe is hydrolysed by the 5 3 exonuclease activity of Taq polymerase. This results in an increase in fluorescence intensity proportional to the accumulation of PCR product. ROX, a rhodamine derivative, is present in the buffer solution as a passive reference label. A normalised signal is generated by the equation: Rn = (Rn + ) (Rn ), where Rn + represents the ratio of reporter signal at any given time and the quencher baseline signal and Rn the ratio of the reporter baseline signal and the quencher baseline signal. Subsequently, the threshold is set at 10 times the standard deviation of the mean baseline emission calculated for the first cycles. The cycle threshold (Ct) is defined as the cycle number at which a sample s Rn fluorescence crossed this threshold, which represents a positive PCR result. Standard curve A standard curve was calculated using linear regression analysis. This standard curve displayed a linear relationship between Ct values and the logarithm of the initial number of positive cells or input plasmid copy number. The dynamic range of the standard curve spanned at least six orders of magnitude. The amount of product in a particular sample is determined by interpolation from a standard curve of Ct values generated from the plasmid dilution series. Results The sensitivity of the CK19 assay as determined by plasmid quantification was very high, with detection of copy numbers less than Table 1. Results of real-time quantitative PCR in peripheral blood of stage IV breast cancer patients (n = 14) and healthy volunteers (n = 23) with plasmid quantification for CK19 CK19 plasmid copies Healthy volunteers Stage IV BC patients (n = 23) (n = 14) Median (95% CI) 1 (1 3) 11 (1 60) BC, breast cancer; CI, confidence interval. 10. The results obtained for both stage IV breast cancer patients and the healthy volunteers are summarised in Table 1. We detected a median CK19+ copy number of one in the control group [95% confidence interval (CI) 1 3], while a median copy number of 11 (95% CI 1 60) was detected in stage IV breast cancer patients (Mann Whitney, P ). Taking the upper limit of the 95% CI of the control group as the cut-off, 9/14 (64.3%) stage IV breast cancer patients and 5/23 (21.7%) of the volunteers were considered positive for CK19. When comparing the results obtained using cell line dilutions with those obtained using plasmid dilutions, a good correlation was observed (r 2 = 0.98). Equivalent results were obtained for both control genes (β2m and GUS) after normalisation (r 2 = 0.996) and their expression level was similar between the analysed samples (Figure 1). Comparison between the detected amount of CK19 positivity using cell line dilutions in our previous work and plasmid dilu-

3 1243 Figure 1. Representation of the correlation between RT PCR cycle threshold (Ct) values for CK19 (triangles), β2m (diamonds) and GUS (crosses) plasmids. Corr, correlation. Table 2. Statistical overview of the results obtained using plasmid and cell line dilutions Quantification method Mean 95% CI of mean Median 95% CI of median IQR SD Healthy volunteers (n = 23) Stage IV BC patients (n = 14) Plasmid a Cell lines b Plasmid a Cell lines b a Results expressed as CK19 copy number. b Results expressed as number of CK19+ cells/ leucocytes. BC, breast cancer; CI, confidence interval; IQR, interquartile range; SD, standard deviation. tions in the current study in blood samples from the patients and volunteers revealed concordance in 78.5% and 73.9% of patients and volunteers, respectively. Table 2 shows a statistical comparison of the results obtained using either cell lines or plasmid dilutions for quantification. Discussion The good correlation (r 2 = 0.98) between the results obtained using cell line dilutions and those obtained using plasmid dilutions confirmed the validity of the results from our previous work. Hence, there was no need to repeat the analysis of these samples. Plasmid dilutions offer several advantages over dilutions based on cell lines. They are more reliable than cell line dilutions for the quantification of gene expression since variation between batches is minimal, whereas differences in expression levels might occur in cell lines grown at different time points or at different passages. Moreover, different cell lines express different levels of CK19, thus, making it difficult to estimate the significance of a certain expression level. Quantification using plasmids is more accurate since absolute copy numbers can be calculated based on concentration measurements. In addition, plasmid quantification is more reproducible since the variation between different PCR runs is extremely low. Our data showed that both β2m and GUS could be used for normalisation of the PCR results. Recently, Stathopoulou et al. [6] showed that the detection of CK19 mrna-positive cells in the blood had prognostic implications for patients with stage I and II breast cancer, because positivity was associated with a reduced disease-free interval (P = ) and overall survival (P = 0.01). However, the PCR results were not quantitative, which might be necessary for the fine tuning of prognostic information. For chronic myeloid leukaemia it was shown that the actual level of positivity was related to the probability of relapse [7]. Moreover, in view of the significant level of falsepositive results observed in healthy volunteers, non-quantitative PCR data might result in erroneous predictions. In another study, it was shown that it is possible to monitor disease response from PB samples using quantitative RT PCR in patients with metastatic breast cancer [8].

4 1244 Table 3. Overview of the cut-off calculation methods described in various publications and how applying them influences our original results using either cell dilutions or plasmid dilutions for quantification Cut-off model With plasmid dilutions a With cell line dilutions b Cut-off value controls (%) stage IV patients (%) a Results expressed as number of CK19 copies. b Results expressed as number of CK19+ cells/ leucocytes. CI, confidence interval; SD, standard deviation. Cut-off value controls (%) stage IV patients (%) Upper value in controls 38 0/23 (0) 3/14 (21.4) 474 0/23 (0) 5/14 (35.7) [11 13] Mean + 2 SD [14] 23 1/23 (4.3) 4/14 (28.6) 336 1/23 (4.3) 5/14 (35.7) Upper limit of 95% CI of median of controls 3 5/23 (21.7) 9/14 (64.3) 136 5/23 (21.7) 10/14 (71.4) The detection of CK19-positive cells might thus be used as a surrogate marker in identifying patients who are at increased risk of relapse and could be candidates for further systemic adjuvant therapy. To the best of our knowledge, there are no prospective studies in the literature that have considered micrometastases as a factor in the basis of selecting treatment option for solid tumours. A recurrent problem in several papers [11 14] is the definition of a cut-off for positivity. Since the markers used for the detection of circulating tumour cells are generally not tumour-specific, and since background expression is often observed in the control groups, it is of great importance to precisely define the cut-off for positivity. In our study, the upper limit of the 95% CI of the median of control group was used as the cut-off, and thus any sample containing greater than three copies was considered positive. Some studies used the highest value in the control group as the cut-off, and thus any values in the samples above this cut-off were considered positive for circulating tumour cells [9 11]. Shammas et al. [14] used as the cut-off definition the mean + 2 standard deviation of the values of CK19 expression in their control group (healthy volunteers and patients without epithelial cancer). Table 3 shows how applying these different definitions of cutoff on our results (both for quantification with cell line dilutions and with plasmid dilutions) resulted in significantly different numbers of positive samples compared with our original method. We feel that the upper limit of the 95% CI of the median of values obtained for the control group is a better cut-off parameter, because high CK19 expression levels could be found in a few control samples. These outlying values might lead to an unnecessary increase in the cut-off value, thus reducing the sensitivity of the assay. This reflects the importance of precisely defining the cutoff value of CK19 positivity. We conclude that plasmid dilutions are more reliable than cell line dilutions for the quantification of gene expression, and more objective criteria for positivity could be defined based on the characteristics of the standard curve (slope and intercept). A more universally accepted agreement on the definition of the cut-off value for positivity is needed to overcome the discrepancy between different studies. This would allow a better evaluation of the significance of detecting circulating tumour cells by RT PCR within each study and a better comparison between the respective studies. Acknowledgements This work was supported by a grant from Vlaamse Liga tegen Kanker, VLK, Brussel and from the Flemish Foundation for Scientific Research (FWO 06260). M.S.I. was supported by a fellowship awarded by the European Society for Medical Oncology (ESMO ). References 1. Diel IJ, Kaufmann M, Costa SD et al. Micrometastatic breast cancer cells in bone marrow at primary surgery: prognostic value in comparison with nodal status. J Natl Cancer Inst 1996; 88: Braun S, Pantel K, Muller P et al. Cytokeratin-positive cells in the bone marrow and survival of patients with stage I, II or III breast cancer. N Engl J Med 2000; 342: Fields KK, Elfenbein GJ, Trudeau WL et al. Clinical significance of bone marrow metastases as detected using the polymerase chain reaction in patients with breast cancer undergoing high-dose chemotherapy and autologous bone marrow transplantation. J Clin Oncol 1996; 14: Aerts J, Wynendaele W, Paridaens R et al. A real-time quantitative reverse transcriptase polymerase chain reaction (RT PCR) to detect breast carcinoma cells in peripheral blood. Ann Oncol 2001; 12: Saad Ismail M, Wynendaele W, Aerts J et al. Quantification of CK19 mrna in peripheral blood (PB) and bone marrow (BM) from primary operable breast cancer (BC) patients pre- and postoperatively to investigate possible shedding of CK19+ cells during the operation. Proceedings of ECCO-11. Eur J Cancer 2001; 37 (Suppl 6): S Stathopoulou A, Vlachonikolis I, Mavroudis D et al. Molecular detection of cytokeratin-19-positive cells in the peripheral blood of patients with operable breast cancer: evaluation of their prognostic significance. J Clin Oncol 2002; 20: Hochhaus A, Weisser A, La Rosee P et al. Detection and quantification of residual disease in chronic myelogenous leukemia. Leukemia 2000; 14:

5 Smith BM, Slade MJ, English J et al. Response of circulating tumor cells to systemic therapy in patients with metastatic breast cancer: comparison of quantitative polymerase chain reaction and immunocytochemical techniques. J Clin Oncol 2000; 18: Braun S, Pantel K. Clinical significance of occult metastatic cells in bone marrow of breast cancer patients. Oncologist 2001; 6: Hawes D, Munro Neville A, Cote RJ. Occult metastasis. Biomed Pharmacother 2001; 55: Slade MJ, Smith BM, Sinnett HD et al. Quantitative polymerase chain reaction for the detection of micrometastases in patients with breast cancer. J Clin Oncol 1999; 17: Ikeda N, Miyoshi Y, Motomura K et al. Prognostic significance of occult bone marrow micrometastases of breast cancer detected by quantitative polymerase chain reaction for cytokeratin 19 mrna. Jpn J Cancer Res 2000; 91: Bieche I, Nogues C, Paradis V et al. Quantitation of htert gene expression in sporadic breast tumors with a real-time reverse transcriptionpolymerase chain reaction assay. Clin Cancer Res 2000; 6: Shammas FV, Deak E, Nysted A et al. Serial quantitative PCR analysis of bone marrow samples from breast cancer patients to monitor systemic micrometastases. Anticancer Res 2001; 21: