| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
From the Institute of Pathology
*
and the Department of Clinical Biochemistry,
Aarhus University Hospital, Aarhus, Denmark
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
One possible answer to this problem may lie in the archives of formalin-fixed, paraffin-embedded (FFPE) tissue specimens held in histopathology departments throughout the world. These collections already represent an invaluable research resource for studying the molecular basis of disease, making it possible to perform large retrospective studies correlating molecular features with therapeutic response and clinical outcome. In recent years, reliable techniques for the immunohistochemical analysis of gene expression in paraffin sections have been developed that form the basis for numerous assays both in research and in routine diagnosis. Similarly, techniques for extraction and analysis of DNA from FFPE tissues have been optimized allowing a range of molecular genetic studies to be performed on archival and routine diagnostic histopathological material. The ability to study mRNA expression in FFPE tissues, even to a limited degree, would be an important advance, opening up the histopathology archive to molecular profiling and allowing analysis of gene expression at the RNA level in standard diagnostic specimens.
Traditionally, it has been considered impossible to perform RT-PCR on archival tissue, partly because the extracted RNA was extensively degraded and partly because the amounts present were too small to be amplified by conventional means.2 However, improved techniques for extracting RNA from fixed specimens based on the use of proteinase K digestion followed by phenol-chloroform extraction,3, 4 and the application of highly sensitive fluorescence-based real-time RT-PCR procedures5, 6, 7 have shown that it is possible to detect mRNA in FFPE tissue. The unanswered question is whether mRNA analysis in FFPE is sufficiently reliable to allow its routine use. Various factors may affect the results of molecular analysis in FFPE compared with fresh/frozen tissues. However, the extent to which these influence quantitative gene expression analysis has not been systematically addressed. The purpose of this study was to establish protocols for real-time RT-PCR on FFPE tissues, to determine the effect on quantitative gene expression analysis of formalin fixation (including prolonged fixation), and to quantify the influence of a period of autolysis before freezing (as a surrogate for prefixation delay in the diagnostic setting). This work is part of a larger study aimed at developing molecular tools for the detection of melanoma micrometastases in sentinel lymph nodes. Therefore, we chose as a model system to study expression of the melanocyte-associated gene MART-1 and the housekeeping genes ß-actin and ß2-microglobulin (ß2-M) in fresh and FFPE lymph nodes with melanoma metastases.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
. For the experiments with FFPE tissues, an additional four pairs of samples from two of the tumors were fixed in 4% neutral buffered formalin for various fixation times (Figure 1B)
. Samples were of similar size to ensure comparable formalin infiltration. After fixation the tissue samples were processed and embedded in paraffin according to standard procedures.
|
Formalin-Fixed, Paraffin-Embedded Tissue
RNA was extracted from FFPE material using the Paraffin Block, RNA isolation kit (Ambion Inc., Austin, TX) with modifications. Briefly, 20-µm paraffin sections were cut, deparaffinized by incubation in xylene for 20 minutes, centrifuged, and washed three times in 100% ethanol. After each wash, the sample was centrifuged at room temperature. After the final wash, the material was air-dried and then incubated in digestion buffer with proteinase K. The residue was then homogenized and incubated overnight at 55°C. RNA was purified by the addition of RNA extraction buffer. Chloroform was then added, followed by additional incubation and centrifugation. The aqueous phase was removed to fresh tubes and the RNA was precipitated with an equal volume of isopropanol in the presence of linearized acrylamide. Samples were incubated at -20°C for at least 30 minutes, after which RNA was pelleted and washed twice in 75% ethanol, with intervening centrifugation at 4°C. After the final centrifugation, the pellet was air-dried and resuspended in 10 µl of RNA storage solution. To remove genomic DNA, all samples underwent DNase treatment according to the manufacturer. All reagents were from Ambion.
RNA Quantification
The amount of RNA in the samples was measured using the RiboGreen RNA quantitation kit (Molecular Probes Inc., Eugene, OR) and by fluorometry (Fluoroskan Ascent FL; Thermo Labsystems, Helsinki, Finland). On the basis of these calculations, all samples were diluted to 0.1 µg/µl.
Primers
Primers for ß-actin (amplicon size, 99 bp), ß2-M (amplicon size, 85 bp), and nested MART-1 (amplicon size, 497 bp/439 bp) were as published.5, 8, 9
We designed primers for an additional nested PCR for MART-1 with a short amplicon size (136 bp/89 bp). The primers were all intron spanning; their sequences are listed in Table 1
.
|
), 20 U RNase inhibitor, and 50 U MULV reverse transcriptase (Applied Biosystems) in a total volume of 20 µl. RT was performed in a Gene Amp PCR system 9700 thermocycler (Applied Biosystems) at 42°C for 30 minutes followed by 99°C for 5 minutes. Two µl of the resulting cDNA was used immediately for real-time PCR, or the cDNA was stored at -20°C.
Reverse Transcriptase-Polymerase Chain Reaction (RT-PCR)
To have as sensitive a method as possible, detection of MART-1 mRNA was performed as nested RT-PCR. To ensure comparability this was done both for frozen and fixed specimens. Each RT-PCR reaction contained 0.1 µg of RNA in a 1x PCR buffer (Qiagen) supplemented with 1.25 mmol/L MgCl2, 0.1 mmol/L of each deoxyribonucleoside triphosphate (dATP, dTTP, dGTP, dCTP), M1 and M2 primers (concentrations as shown in Table 1
), 1 U Taq-polymerase (Qiagen), 50 U MULV reverse transcriptase (Applied Biosystems), and 20 U RNase inhibitor in a total volume of 40 µl. One step RT-PCR amplification was then performed in a Gene Amp PCR system 9700 thermocycler according to the following schedule: RT at 37°C for 60 minutes followed by denaturation at 94°C for 1 minute. The PCR amplification profile was: 15 seconds at 94°C, 30 seconds at 65°C, and 30 seconds at 72°C. After 30 cycles for MART-1 (long) and 10 cycles for MART-1 (short), the PCR product was extended at 72°C for 7 minutes. After amplification, 2 µl of the PCR product was used as template for the second round PCR, which was performed as real-time PCR (see below).
Real-Time PCR
Two µl of cDNA (ß-actin, ß2-M,) or 2 µl of PCR product (MART-1) were used as template for real-time PCR in a reaction mixture containing 4 mmol/L of MgCl2, primers as shown in Table 1
, 2 µl mix from SYBR Green I kit [containing TaqDNA polymerase, reaction buffer, deoxyribonucleoside triphosphate (dATP, dTTP, dGTP, dCTP), SYBR Green I dye, and 10 mmol/L of MgCl2; Roche Molecular Biochemicals, Indianapolis, IN], the volume being adjusted to 20 µl with nuclease-free water. The samples were amplified in the Lightcycler System (Roche) and the PCR was performed by an initial denaturation step at 95°C for 30 seconds and then 40 cycles with a 95°C denaturation immediately followed by annealing (temperature given in Table 1
) for 5 seconds and 72°C extension for 10 seconds. After 40 cycles a melting curve was generated for the final PCR product of all genes investigated by decreasing the temperature to 65°C for 10 seconds followed by a slow increase in temperature to 95°C. During the slow heating process the fluorescence is measured at 0.2°C increments. To further ensure specificity, a gel electrophoresis was conducted for selected samples of each specific product. It consistently revealed bands of the expected length.
Experimental Design
RT was performed in duplicate for each RNA sample. RNA for a calibration curve was included in every RT together with a high- and a low-positive control in duplicate and one negative control (without RNA) to exclude potential contamination. PCR reactions for both standard curve (including positive and negative controls) and samples were performed from each RT. The interassay and intra-assay variation was determined by analyzing the positive controls in 10 consecutive analytical runs (interassay variation) with each sample analyzed in duplicate (intra-assay variation) (Table 2)
.
|
Standard Curve
Construction of a standard curve for ß-actin and ß2-M was achieved using RNA from a cultured urothelial cell line HCV 29.10
The RNA was diluted in water to give five different calibrators with known amounts of HCV 29 RNA in 10-fold dilutions (calibrators 1 to 5 contained 1.0 µg to 0.00001 µg of HCV29 RNA/µl; 1 µl of calibrator RNA was used in each calibrator RT-PCR). The amount of housekeeping gene mRNA present in 1 µg of HCV29 RNA was defined as 1 arbitrary unit of HCV29 ß-actin and ß2-M mRNA, respectively. Results of unknown samples were expressed as arbitrary units of HCV29 RNA/µg sample RNA. For MART-1 a standard curve of eight calibrators was obtained using RNA isolated from the melanoma cell line HTB 72 (American Type Culture Collection, Manassas, VA). The calibration curve was obtained using dilutions of RNA isolated from 400 HTB 72 cells (calibrators 1 to 8 contained RNA ranging from 400 to 0.16 HTB 72 cells/µl; 1 µl of calibrator was used in each calibrator RT-PCR). The amount of MART-1 mRNA present in 400 HTB 72 cells was defined as 1 arbitrary unit of HTB 72 cells. Results of unknown samples are expressed in arbitrary units of HTB 72 cells/µg sample RNA.
For each amplicon investigated a corresponding standard curve is generated with the same primers (both short and long) as the investigated samples and included in each run.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
. This depended on the amount of mRNA present. For high quantity controls, the analytical interassay imprecision (coefficient of variation) ranged from 12 to 26%, whereas the range for low quantity controls was 31 to 40%. In most of the experiments in our study the samples were in the range of the high quantity controls.
Effect of Freezing Delay on mRNA Levels (Figures 2 and 3)
In the first series of experiments, we investigated the effect of delayed freezing (autolysis) on the quantification of ß-actin, ß2-M, and MART-1 RNA; amplicons were of comparable size (99 bp, 85 bp, and 136 bp, respectively).
|
|
. For both genes it appeared that the decrease in mRNA was most pronounced within the first hours of autolysis (Figure 2)
.
Cell-Specific Gene
As expected for a melanocyte-specific gene MART-1, the level of gene expression showed marked variation comparing samples from different tumors (data not shown). Also as expected from the housekeeping gene results, variation was found depending on freezing delay. However, expressing MART-1 relative to the content of ß-actin markedly diminished the observed variation (Figure 3)
. Comparable results were obtained if ß2-M was used as the reference gene (data not shown). In general, there was no apparent change in relative expression of MART-1 in samples after between 0 and 24 hours of autolysis.
Effect of Formalin Fixation, Fixation Time, and Amplicon Size on mRNA Quantification (Figures 4 and 5)
Attention was focused on RNA integrity and yield as a function of fixation time using expression of ß-actin and MART-1 mRNAs as a model. In all of the samples, mRNA specific for ß-actin and MART-1 could be detected in FFPE melanoma tissue, even after prolonged formalin fixation for up to 21 days (Figures 4 II and 5 II)
. However, a marked decrease (85 to 99%) in mRNA levels for both ß-actin and MART-1 was seen in FFPE tissue (even after 1 day of fixation) compared with matched frozen tissue. This decrease was most pronounced when amplifying long fragments.
|
|
, whereas an up to 10-fold difference was seen for MART-1 (amplicon size, 497 bp versus 136 bp; Figure 5 II
). In the fixed tissues the mRNA recovery, using short compared with long amplicons, was 20-fold greater for ß-actin and 100-fold greater for MART-1. Fixation for prolonged periods (5 days or more) did not show any systematic decrease in PCR product for neither short nor longer amplicon sizes (Figures 4 and 5)
. |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
More recently several groups have reported that the small amounts of RNA available in FFPE tissues can be amplified using highly sensitive real-time RT-PCR techniques.5, 7, 13 Our study confirms these findings. Using real-time RT-PCR, we were consistently able to detect mRNA for both MART-1 and for the housekeeping genes ß-actin and for ß2-M in FFPE melanoma metastases, even after prolonged fixation in formalin for up to 3 weeks. This was partly dependent on the use of optimized extraction procedures with proteinase K digestion. Clearly, formalin fixation markedly reduces the amount of RNA that can be extracted from paraffin blocks—in our study by up to 99% when using standard primers. However, by using PCR primers that amplify short sequences it was possible to increase the yield of amplifiable RNA by up to 100-fold. The increased sensitivity we found using smaller amplicon sizes is consistent with other studies that have suggested that formalin fixation degrades RNA.5 Although, length of fixation may be a problem for quantification, our experiments did not show any systematic decrease in PCR product as a function of fixation length.
One important factor that may limit the ability to quantify mRNA in FFPE tissues is delay in fixation. In particular, it has been widely assumed that because prefixation times vary for routine specimens, and will rarely be known retrospectively, that quantification of gene expression in archival specimens would be unreliable.14 However, although much is known about the regulation and degradation of mRNA in experimental systems,15, 16 surprisingly little has been published concerning the stability and turnover of mRNAs in clinical specimens. Given that typical half-lives for mRNAs vary from a few minutes for the most labile to more than 24 hours for the more stable gene products such as those from globin genes, it would be expected that delay in fixation would result in a skewed representation of mRNAs extracted from paraffin blocks. Most previous studies of RNA decay in in vivo specimens have been performed on frozen tissues from autopsy brains.17, 18, 19, 20 These show surprisingly stable retention of a variety of mRNAs postmortem, although by their nature they lack evaluation of RNA degradation in the important first hours of death. Other studies have looked at the effect of RNA degradation on quantification of gene expression by RT-PCR in cultured cells.21, 22 However, for several reasons it is uncertain whether the results of such studies can be directly extrapolated to the in vivo situation. In particular, gene expression patterns and rates of RNA degradation may not be representative in in vitro cells separated from other tissue elements. Thus, it is important to study the factors affecting RNA degradation in clinical specimens.
To evaluate in vivo RNA decay we chose a model reflecting daily routine as closely as possible. Freezing delays up to 24 hours were chosen to reflect the absolute worst case situation for delayed tissue processing. For both ß-actin and ß2-M genes a decrease in mRNA level was seen, most of this decrease apparently occurring within the first 2 hours of autolysis after which a steady state appeared. Godfrey and colleagues5 have also looked at the effect on quantification of several mRNAs of delayed freezing/fixation. They studied liver specimens from a single patient, and found that a delay of up to 12 hours before fixation did not change the relative expression of the genes studied. Our results confirm this finding and further show that this applies for tissues other than RNA-rich liver, and when a larger series of patient tumors is studied with parallel analyses. There are several possible explanations for the relative stability of mRNA levels in tissues after surgical removal. Firstly, although there will be progressive depletion of oxygen and nutrients, it is reasonable to assume that initially the tissue will continue to metabolize and to transcribe RNA. Secondly, and probably more importantly, tissue and cell integrity will be primarily maintained during the autolytic period limiting the access of RNases to intracellular mRNA. Indeed, studies of autopsy brains show that the relative stability of RNA postmortem is lost once the tissue has been frozen and then thawed out, after which RNA decay is greatly accelerated.18, 20 Freeze/thaw cycles will not only stop tissue metabolism, but will also rupture cells, making the intracellular compartment more accessible to RNases.
Our study provides useful information about the two housekeeping genes investigated. The choice of such genes is important, but often difficult. The ideal housekeeping gene is one that is expressed at constant, moderate levels in all tissues. When used as a reference gene it must reflect the target gene with respect to turnover rate. This is of course impossible to achieve and in practice these genes may show variation in expression in different tissues and under different experimental conditions.22
Furthermore, little is known about the expression of housekeeping genes in tumor tissues, although some have been shown to be up-regulated in cancer.23, 24
In our experiments, the higher SEM for detecting ß2-M mRNA compared with ß-actin mRNA is more likely to be related to analytical imprecision (Table 2)
than to biological variation. Thus, there is no evidence that intertumor variation in expression of these two housekeeping genes is likely to be a significant source of imprecision when they are used as controls to quantify mRNA levels in melanoma.
We also studied expression of a cell-specific mRNA—MART-1. This gene product is a transmembrane molecule found on melanocytic cells. Neither the biological function of MART-1, nor the half-life of its mRNA are known. We wanted to see if there was any significant decrease in expression of this gene as a function of autolysis. The absolute quantification of MART-1 mRNA showed considerable differences comparing samples, both intratumoral and marked intertumoral variation being found (data not shown). However, to correct for variation because of insufficient starting material and inefficient cDNA synthesis, we chose to express MART-1 mRNA levels relative to a reference gene. MART-1 mRNA was normalized to both ß2-M and ß-actin mRNAs to evaluate which of the two housekeeping gene products it resembled most with respect to turnover rate. This analysis gave almost identical results whether MART-1 was expressed relative to ß-actin or to ß2-M (data not shown in full). MART-1 mRNA showed relatively constant expression with only minor intratumoral variation, whereas the intertumoral (presumably biological) variation was still considerable (Figure 3)
. Thus, the turnover rate for MART-1 appears to reflect both ß-actin and ß2-M, making them both suitable reference genes in our experiments. MART-1 mRNA could be easily detected, even after 24 hours of prefreezing delay. No significant change in expression levels could be identified as a function of autolysis time. This result suggests that mRNA may be more robust in vivo than sometimes supposed, although this should be interpreted with care. An alternative explanation may be that minor degrees of change in mRNA levels are masked by tumor heterogeneity. Tumors are composed of a complex mix of neoplastic cells including various subclones. This intratumoral heterogeneity is added to by the presence of a variable component of nonneoplastic stromal and inflammatory cells, by areas of dead and dying cells, by variation in the level of nutrients and oxygen, and by variation in the degree of vascularization and in the levels of cytokines and other active biomolecules. All these factors have the potential to affect the level of expression of particular genes. Thus, tumor heterogeneity may be a source of possibly serious bias when quantifying RNA extracted from tissues.25
The effects of tumor heterogeneity can be limited by analyzing several different areas of a tumor and by selecting tissue samples under microscopic control.
The improved PCR yield for MART-1 mRNA achieved by amplifying short sequences was seen not only in the fixed samples, but also when analyzing fresh-frozen tissue (Figure 5)
. This may partly reflect a primer-related change in PCR efficiency. Alternatively, the longer amplicon size (497 bp) used initially for MART on frozen samples compared with ß-actin (313 bp), may be more vulnerable to even minor degrees of degradation occurring within the first few minutes after surgical removal. The reliability of the molecular results obtained from real-time RT-PCR is dependent on the analytical precision of the detection methods used, but only a few studies have addressed this important point. We have determined the precision of the real-time RT-PCR method used for controls in the measuring range of our samples as well as in controls with mRNA levels 
100-fold lower. In the range of the samples we find an acceptable imprecision for mRNA from both the ß-actin and ß2-M housekeeping genes with intra- and interassay variations of 10% and 20%, respectively. Even for MART-1 mRNA, which is quantified in a nested RT-PCR design there is an acceptable imprecision with 10% intra-assay and 26% interassay variation, levels comparable to the two household genes examined that support the contention, that the data for the nested products are quantitative. As expected, lowering the mRNA content of the controls 100-fold below the values of our samples resulted in an increase in the interassay imprecision to 30% for the household genes and 40% for MART-1.
Our findings confirm that clinical tumor specimens removed by routine surgical procedures and snap-frozen can be safely used for studies of gene expression. Although we cannot exclude the possibility that some particularly labile mRNAs may be preferentially lost immediately after removal, quantification of more stable transcripts relative to housekeeping gene mRNAs would seem to be possible even with relatively prolonged delays up to 24 hours before freezing or fixation of the tissue. Furthermore, by using highly sensitive real-time RT-PCR with amplification of short target sequences, mRNA for a variety of genes can be quantified in FFPE tissues. An important proviso to remember is that the amount of RNA that can be extracted from FFPE tissue represents only a minor part of that which is obtainable from fresh-frozen tissue and that the analytical imprecision is increased when the amount of amplifiable RNA available is small. Although variations in the time before tissue processing seem to be less important sources of imprecision, than previously assumed, variations in length of fixation may still be a problem for quantification. Normalizing results to reference genes seems to be mandatory for the reliability of the molecular results. Our model demonstrates a way to select the appropriate reference gene. Although it will be important to confirm these results in studies on other genes than those we have looked at, our findings open up the possibility of performing retrospective molecular profiling studies of archive tissues at the RNA level and using these techniques in a routine diagnostic setting.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Supported by the Clinical Research Unit of the Danish Cancer Society; the Institute for Experimental Clinical Research, Aarhus University; the Danish Medical Society, Copenhagen; Fritz, Georg, and Marie Cecilie Glud’s Foundation; Max and Inge Wørzners Memorial Foundation; the Danish Medical Association, Højmosegårds Legate; and Forskningsinitiativet, County of Aarhus.
Accepted for publication November 4, 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  J.-P. Brosseau, J.-F. Lucier, E. Lapointe, M. Durand, D. Gendron, J. Gervais-Bird, K. Tremblay, J.-P. Perreault, and S. A. Elela High-throughput quantification of splicing isoforms RNA, February 1, 2010; 16(2): 442 - 449. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 U Siebolts, H Varnholt, U Drebber, H-P Dienes, C Wickenhauser, and M Odenthal Tissues from routine pathology archives are suitable for microRNA analyses by quantitative PCR J. Clin. Pathol., January 1, 2009; 62(1): 84 - 88. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. R. Varadhachary, D. Talantov, M. N. Raber, C. Meng, K. R. Hess, T. Jatkoe, R. Lenzi, D. R. Spigel, Y. Wang, F. A. Greco, et al. Molecular Profiling of Carcinoma of Unknown Primary and Correlation With Clinical Evaluation J. Clin. Oncol., September 20, 2008; 26(27): 4442 - 4448. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Doleshal, A. A. Magotra, B. Choudhury, B. D. Cannon, E. Labourier, and A. E. Szafranska Evaluation and Validation of Total RNA Extraction Methods for MicroRNA Expression Analyses in Formalin-Fixed, Paraffin-Embedded Tissues J. Mol. Diagn., May 1, 2008; 10(3): 203 - 211. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. S. Knudsen, A. N. Allen, D. F. McLerran, R. L. Vessella, J. Karademos, J. E. Davies, B. Maqsodi, G. K. McMaster, and A. R. Kristal Evaluation of the Branched-Chain DNA Assay for Measurement of RNA in Formalin-Fixed Tissues J. Mol. Diagn., March 1, 2008; 10(2): 169 - 176. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Kidd, B. Nadler, S. Mane, G. Eick, M. Malfertheiner, M. Champaneria, R. Pfragner, and I. Modlin GeneChip, geNorm, and gastrointestinal tumors: novel reference genes for real-time PCR Physiol Genomics, August 20, 2007; 30(3): 363 - 370. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. Loudig, E. Milova, M. Brandwein-Gensler, A. Massimi, T. J. Belbin, G. Childs, R. H. Singer, T. Rohan, and M. B. Prystowsky Molecular restoration of archived transcriptional profiles by complementary-template reverse-transcription (CT-RT) Nucleic Acids Res., August 1, 2007; 35(15): e94 - e94. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Sircar, L. Gaboury, L. Ouadi, M. Mecteau, E. Scarlata, F. Saad, A. Aprikian, S. Tanguay, S. Lapointe, C. Lussier, et al. Isolation of human prostatic epithelial plasma membranes for proteomics using mirror image tissue banking of radical prostatectomy specimens. Clin. Cancer Res., July 15, 2006; 12(14): 4178 - 4184. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Talantov, J. Baden, T. Jatkoe, K. Hahn, J. Yu, Y. Rajpurohit, Y. Jiang, C. Choi, J. S. Ross, D. Atkins, et al. A Quantitative Reverse Transcriptase-Polymerase Chain Reaction Assay to Identify Metastatic Carcinoma Tissue of Origin J. Mol. Diagn., July 1, 2006; 8(3): 320 - 329. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. W. Tothill, A. Kowalczyk, D. Rischin, A. Bousioutas, I. Haviv, R. K. van Laar, P. M. Waring, J. Zalcberg, R. Ward, A. V. Biankin, et al. An Expression-Based Site of Origin Diagnostic Method Designed for Clinical Application to Cancer of Unknown Origin Cancer Res., May 15, 2005; 65(10): 4031 - 4040. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Bibikova, D. Talantov, E. Chudin, J. M. Yeakley, J. Chen, D. Doucet, E. Wickham, D. Atkins, D. Barker, M. Chee, et al. Quantitative Gene Expression Profiling in Formalin-Fixed, Paraffin-Embedded Tissues Using Universal Bead Arrays Am. J. Pathol., November 1, 2004; 165(5): 1799 - 1807. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. M. Gjerdrum, B. S. Sorensen, E. Kjeldsen, F. B. Sorensen, E. Nexo, and S. Hamilton-Dutoit Real-Time Quantitative PCR of Microdissected Paraffin-Embedded Breast Carcinoma: An Alternative Method for HER-2/neu Analysis J. Mol. Diagn., February 1, 2004; 6(1): 42 - 51. [Abstract] [Full Text]
|
|
|
|
|
|
M. Cronin, M. Pho, D. Dutta, J. C. Stephans, S. Shak, M. C. Kiefer, J. M. Esteban, and J. B. Baker Measurement of Gene Expression in Archival Paraffin-Embedded Tissues: Development and Performance of a 92-Gene Reverse Transcriptase-Polymerase Chain Reaction Assay Am. J. Pathol., January 1, 2004; 164(1): 35 - 42. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
