| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Division of Clinical Chemistry, Department of Biomedicine and Surgery, Linköping University, S-581 85 Linköping, Sweden.
2 Childhood Cancer Research Unit, Department of Woman and Child Health, Karolinska Institutet, Karolinska Hospital, S-171 76 Stockholm, Sweden.
3 Department of Clinical Chemistry, University Hospital, S-581 85 Linköping, Sweden.
aAuthor for correspondence. Fax 46-13-22-32-40; e-mail bertil.kagedal{at}lio.se.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsDiscussionReferences |
|---|
Methods: We used real-time reverse transcription-PCR. The calibrator was constructed from a segment of tyrosine hydroxylase mRNA that included the target. Blood and bone marrow samples from 24 children with neuroblastoma and 1 child with ganglioneuroma were analyzed. Controls were blood samples from the cords of 40 babies, from 58 children 6 months to 15 years of age, and from 34 healthy adults, as well as from 12 children with other diseases.
Results: The detection limit was 
70 transcripts/mL. All 144 blood controls were below this limit. At diagnosis, blood tyrosine hydroxylase mRNA was higher in children with widespread disease (stage 4/4S; n = 6; range, 203–46 000 transcripts/mL) than in patients with localized disease (stages 1–3; n = 6; 
83 transcripts/mL; P = 0.002). Bone marrow from all five children with localized disease had concentrations <72 transcripts/mL, whereas five of six stage 4 patients had increased concentrations (6000–8 000 000 transcripts/mL; P <0.05). In nine children in whom tyrosine hydroxylase mRNA was measured repeatedly, the results corresponded to the clinical course.
Conclusion: Quantitative analysis of tyrosine hydroxylase mRNA in blood and bone marrow is reliable and easy to perform and may be used for upfront staging, prognostic assessment, and treatment monitoring of neuroblastoma.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsDiscussionReferences |
|---|
Neuroblastoma patients are divided into five stages, 1–4 and 4S, according to the International Neuroblastoma Staging System. Despite multimodal therapy, children with metastatic stage 4 disease have a poor prognosis (2). Neuroblastoma cells frequently infiltrate bone marrow; therefore, evaluation of bone marrow is routinely performed as a component of clinical staging (3)(4).
Solid tumors continuously shed cells into the circulation (5). In neuroblastoma, immunocytochemical studies have demonstrated that tumor cells can be detected in peripheral blood from affected children (6). More recently, molecular methods have been developed for the detection of small amounts of neuroblastoma cells. These methods use reverse transcription of tyrosine hydroxylase mRNA followed by amplification of a specific product by PCR. The methods are suitable for qualitative detection of minimal residual disease (MRD)1 in neuroblastoma but are hampered by their inability to quantify disease (7)(8)(9)(10). To improve detection of MRD and evaluation of therapy, quantitative methods are required (3). We therefore investigated the possible use of real-time PCR (11) for this purpose. In many aspects, the method follows the principles listed in a recent publication on the analysis of malignant melanoma transcripts (12).
The aim of the present study was to develop a quantitative reverse transcription-PCR (RT-PCR) method for tyrosine hydroxylase mRNA for the detection of neuroblastoma cells in blood and bone marrow at diagnosis and during therapy and to evaluate the usefulness of this method for early and sensitive detection of relapsing disease during follow-up. This method may provide useful information for early sensitive evaluation of treatment response in children with neuroblastoma.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsDiscussionReferences |
|---|
reagents
The primers used to amplify tyrosine hydroxylase were designed to be mRNA-specific. The forward primer (FP) was placed in exon 5 and 6 spanning an intron, and the reverse primer (RP) was placed in exon 6 and 7, also spanning an intron (Table 1
). The length of the amplicon was 86 bp. The probe was labeled with 6-carboxyfluorescein (FAM) at the 5' end and 6-carboxytetramethylrhodamine (TAMRA) at the 3' end. The primers and the fluorogenic probe were designed using the program Primer Express (PE Applied Biosystems).
|
All oligonucleotides were from Scandinavian Gene Synthesis. The sequences of the primers and probe are shown in Table 1
.
calibrators
RNA was extracted from IMR-32 cells as described below. cDNA was then synthesized by PCR using two primers (Table 1
) selected to give an amplicon including the target. This fragment was expected to contain 299 bp. The corresponding band on gel electrophoresis was purified with the GFXTM PCR DNA and Gel Band Purification Kit (Amersham Pharmacia), and the sequence was verified by sequence analysis. From the absorbance values, this calibrator was calculated to contain 2.064 x 1014 molecules/mL. The solution was first diluted to contain 1 x 106 molecules/µL. Using RNase-free water, we prepared a diluent containing 0.5 mL/L Tween 20 and 10 mg/L tRNA (Sigma) as a carrier. Using this diluent, we prepared a solution containing 1 transcript/2.5 µL (volume taken for the PCR step) and analyzed this solution 80 times for tyrosine hydroxylase mRNA. The proportion of negative results was recorded. From the Poisson distribution (13), the concentration was verified to be 0.935 transcripts/2.5 µL. A final stock calibrator solution containing 106 transcripts/µL was then made from an earlier stock of the preparation and contained 0.5 mL/L Tween 20 and 10 mg/L tRNA (Sigma). We divided the calibrator solution into 15-µL portions and stored them at -70 °C. We analyzed 10-fold serial dilutions of the calibrator to generate a calibration curve (Fig. 1
).
|
analytical control samples
Preparation of cDNA controls from neuroblastoma cells.
IMR-32 cells were cultured, the RNA was extracted, and the cDNA was synthesized according to the method described below. The cDNA was then diluted to three different concentrations with RNase-free water, aliquoted in microtubes, and stored at -70 °C.
Preparation of RNA and cDNA controls from neuroblastoma cells added to blood.
IMR-32 cells were cultured and added to blood from a healthy individual. Both the RNA extraction and cDNA synthesis procedures were scaled up to give sufficient material for long-term storage and analysis. RNA was extracted and pooled. Part of the pool was used for cDNA synthesis, which was performed with the same proportions between RNA sample and reagents as described below. The pools were aliquoted (10 µL) into microtubes and stored at -70 °C.
blood and bone marrow collection and rna extraction
Blood was collected in 10-mL tubes (Becton Dickinson) containing 1.5 mL of acid-citrate-dextrose (ACD) or other tubes containing ACD solution to give the same proportion between ACD and blood. Bone marrow was taken from the posterior iliac crest during anesthesia and collected in a syringe. A measured volume was transferred to ACD solution and taken to the laboratory. We extracted total RNA from 1.5 mL of blood (in duplicates when possible) and bone marrow, using the RNeasy Blood Mini Kit (Qiagen GmbH). The erythrocytes were lysed with EL buffer, and the leukocytes were washed and pelleted twice by centrifugation. The leukocytes were then lysed with 600 µL of RLT buffer containing mercaptoethanol and homogenized with a QIAshredder. After the addition of one volume (600 µL) of 700 mL/L ethanol, the homogenate was applied to an RNeasy mini spin column. The column was washed once with RW1 buffer and twice with RPE buffer. Finally, the RNA was eluted with 40 µL of RNase-free water.
first-strand CDNA synthesis
The RNA (10 µL) was denatured at 70 °C for 5 min and then placed on ice. An equal volume (10 µL) of cDNA mixture was added to give the following final reagent concentrations: 1x First Strand Buffer (Invitrogen), 7.5 mmol/L dithiothreitol (Invitrogen), 500 µmol/L deoxynucleotide triphosphates (Pharmacia), 50 µmol/L random hexamer (Pharmacia), 1 unit/µL RNasin (Promega), and 10 U/µL murine Moloney leukemia virus reverse transcriptase (Invitrogen). First-strand cDNA synthesis was performed at 40 °C for 45 min (Perkin-Elmer 9700 instrument) followed by inactivation of the enzyme by heating for 5 min at 95 °C.
pcr
From the first-strand cDNA, 2.5 µL was taken for amplification in a total volume of 25 µL. The final reagent concentrations were as follows: 1x TaqMan PCR Buffer (Perkin-Elmer), 5 mM MgCl2 (Perkin-Elmer), 200 µM deoxynucleotide triphosphates, 100 nM each of forward and reverse primer and tyrosine hydroxylase probe, and 0.05 U/µL AmpliTaq Gold polymerase (Perkin-Elmer). The samples were amplified in an ABI Prism 7700 Sequence Detector System (Perkin-Elmer). The PCR profile used was 95 °C for 10 min followed by 40 cycles of 15 s at 95 °C and 1 min at 60 °C. For PCR, both of the calibrators and samples were run in triplicate.
patients
The children enrolled in the study were those children treated at Linköping University Hospital, Umeå University Hospital, and the Astrid Lindgren Children’s Hospital in Stockholm during the period January 1999 to April 2001. The Ethics Committees of the involved centers approved this multicenter study, and informed consent was obtained from the parents.
A total of 25 children (24 with neuroblastoma and 1 with ganglioneuroma) were investigated. An overview of their clinical stages is given in Table 2
together with information on established risk factors. The diagnosis was obtained according to the international agreement of diagnostic requirements for neuroblastoma (4). The children were between 2 months and 6 years of age at the time of diagnosis and either had evidence of disease when the study started or were diagnosed during the time of the study. All patients were treated according to international protocols in a risk-based fashion depending on age, stage, and biological risk factors. In short, children with locoregional tumors (stages 1–3 in the International Neuroblastoma Staging System) without MYCN amplification were treated with surgery and received no adjuvant chemotherapy, whereas children with MYCN-amplified tumors and those >1 year with stage 4 received intensive multimodal therapy with chemotherapy, surgery, stem cell transplantation, local irradiation, and 13-cis-retinoic acid. Infants with stage 4S and stage 4 tumors without MYCN amplification received little or no therapy except surgery. At the follow-up visits, blood and bone marrow samples were collected for analysis of tyrosine hydroxylase transcript; thus there was an interindividual variation in the number of analyses.
|
controls
We analyzed cord blood from 40 newborns, blood from 58 healthy children of different ages (6–12 months, n = 19; 5–9 years, n = 22; 11–16 years, n = 17), and blood from 34 healthy adult blood donors. Blood (n = 12) and bone marrow (n = 19) samples from children with diseases other than neuroblastoma, predominantly acute lymphoblastic leukemia, were also investigated. The Ethics Committee in Linköping also approved the use of blood from control individuals, and informed consent was obtained from the parents or, in appropriate cases, from the children themselves.
calculations and statistics
For calculation of the blood or bone marrow concentrations of tyrosine hydroxylase mRNA, the transcript concentration read from the calibration curve (transcript/µL) was multiplied by the factor 53. This value can be derived as follows: During RNA purification the sample is concentrated by a factor 37.5 (1500 µL/40 µL), and during cDNA synthesis the sample is diluted twofold. The concentration read from the calibration curve thus should be multiplied by the factor 0.053 (2/37.5) to give the sample concentration in transcripts/µL and by 53 to give the concentration in transcripts/mL.
The results from series of duplicates and triplicates were stratified into different groups (ranges) according to the transcript concentration in the sample, and the imprecision was calculated by ANOVA using the program SPSS. The mean SD of each group was defined as the square root of the within-sample variance.
Differences between groups were tested by the Wilcoxon rank-sum test and/or two-sided Fisher exact test. The sign test was used for comparison between samples obtained during surgery and those obtained pre- and postoperatively.
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsDiscussionReferences |
|---|
calibration curves
We used 10-fold dilutions of the stock calibrator solution, which contained 106 molecules of the calibrator amplicon, to generate the calibration curves. The logarithm of the dilution gave a linear curve (Fig. 1
) when plotted against the Ct. Forty-five series were run within a period of 20 months. We observed no changes in Ct with time when the same threshold was used (Fig. 2
). To estimate the variability of the PCR, we calculated the number of transcripts in each calibrator from the calibration curve. We calculated a CV from the triplicates of the 45 series. For each calibrator, the mean number of cDNA transcripts/µL is given in parentheses after the mean CV. The mean CVs were 6.1% (0.85 x 106 transcripts/µL), 8.7% (1.1 x 105 transcripts/µL), 9.8% (1.1 x 104 transcripts/µL), 14% (1.1 x 103 transcripts/µL), 24% (1.1 x 102 transcripts/µL), 42% (0.94 x 101 transcripts/µL), and 121% (1.05 x 100 transcripts/µL).
|
imprecision and detection limit
To evaluate imprecision we prepared cDNA control samples from cultured IMR-32 cells and stored these samples aliquoted at -70 °C. These cDNA samples were included in 10 of the above-mentioned analytical series. The Cts for these controls were also unchanged with time. These control samples were diluted to contain amounts corresponding to 666 000, 6660, and 67 transcripts/mL of blood. The results were as follows [mean ± SD (CV)]: 690 500 ± 144 000 transcripts/mL (21%), 8620 ± 2244 transcripts/mL (26%), and 69.3 ± 32 transcripts/mL (46%).
To estimate the long-term imprecision we prepared one pool of RNA and one pool of cDNA from blood with added cultured IMR-32 cells. These controls were analyzed in the same analytical series as blood samples for a period of 11 months (n = 30 for cDNA and n = 27 for RNA), and the results corresponding to blood sample concentrations of transcripts were calculated. The mean ± SD (CV) for cDNA was 105 000 ± 20 000 transcripts/mL (19%), and that for RNA was 168 000 ± 47 000 transcripts/mL (28%).
For analysis of patients’ samples we performed the PCR step in triplicates. Using the data collected over several months, we calculated the imprecision at different transcript concentrations in blood. As shown in Table 3
, there was a concentration-dependent variation of the mean CV similar to that observed with the calibrators. For the 17 samples with the lowest concentrations (50–100 transcripts/mL), the mean SD was 36 transcripts/mL. Usually, 2 SD from zero is used for defining the lowest detectable concentration. This approach yielded a detection limit of 72 transcript/mL.
|
We also evaluated the imprecision for the whole procedure. Duplicate samples were taken from the ACD blood (from neuroblastoma patients) and processed according to the above procedure (including lysis of cells, extraction of RNA, first-strand cDNA synthesis, and PCR; the PCR step was run in triplicate for each of the two extracts as described in Materials and Methods). The withinseries imprecision was then calculated from the duplicates values (Table 4
). The mean CV seemed to be independent of the transcript concentrations in the samples. For the six samples with the lowest concentrations, the mean SD was 22 transcripts/mL. Calculated as above, this would give a detection limit of 44 transcripts/mL. When additional samples with numerical values <50 transcripts/mL were included, the SD was 20 transcripts/mL and the detection limit was 40 transcripts/mL.
|
As a prerequisite for evaluation of recovery studies using neuroblastoma cells added to blood we needed to know whether inhomogeneity of the cells used for the studies would add to the variability of final results. We therefore added neuroblastoma cells (CHP-212) and lysed neuroblastoma cells to blood and blood lysates, respectively, for replicate analysis. The data obtained were used to calculate which components of the procedure added most to the variability of the results. As can be seen from Table 5
, the PCR step added less to the total variation than did the steps before. This was most obvious in the samples with high transcript concentrations (35 000 transcript/mL of blood) compared with the samples with lower transcript concentrations (3500 transcripts/mL of blood). The total variability seemed not to be affected by inhomogeneity of the added cells.
|
recovery
To evaluate the comparability between sample and calibrator cDNA, we extracted mRNA from blood samples containing tyrosine hydroxylase mRNA, synthesized cDNA as described, and prepared a pool of the cDNA. The pool was used as obtained and diluted 1:5 and 1:10. To nine parts of the samples, we added one part of the calibrator to increase the concentration by 105 transcripts/mL. The added amount of cDNA was also measured in the same series and was 1.15 x 105 transcripts/mL. The measured recoveries of cDNA calibrator added to sample cDNA were quantitative (Table 6
).
|
Cultured IMR-32 cells were added to blood samples collected in ACD solution (n = 3) and to phosphate-buffered saline (n = 3) and analyzed for tyrosine hydroxylase. The cell concentrations were 104 and 105 cells/mL. The recoveries in blood were 57% ± 15% and 52% ± 16%, respectively, when the number of transcripts from cells added to phosphate-buffered saline was taken as 100%.
stability of blood samples
IMR-32 cells were added to blood samples to give 104 cells/mL. The samples stored at room temperature were stable for 48 h.
controls
For control purposes, we analyzed blood from the cords of 40 babies, from 58 children 6 months to 15 years of age, and from 34 healthy adult blood donors, as well as from 12 children with diseases other than neuroblastoma. Tyrosine hydroxylase mRNA in these controls was below the detection limit. However, in leukemic bone marrow, increased amounts were found in 5 of 19 cases. The range for these five results was 165-3000 transcripts/mL.
tyrosine hydroxylase MRNA in blood and bone marrow at diagnosis
A total of 16 children received their diagnosis during the study, and for 12 of them, blood samples had been obtained at diagnosis before the start of therapy (Table 7A
). In children with locoregional disease (stages 1–3; n = 6), the blood tyrosine hydroxylase mRNA concentration was 83 transcripts/mL (cases 2 and 4–8; Table 1
of the Data Supplement that accompanies the online version of this article at http://www.clinchem.org/content/vol49/issue1/), and in children with widespread disease (stage 4; n = 6), the concentration range was 203–46 000 transcripts/mL (cases 12, 14, and 19–22; Table 1
in the Data Supplement). This difference between the groups was significant (P <0.01).
|
Tyrosine hydroxylase mRNA could not be detected in bone marrow samples (Table 7B
) obtained from the five children with locoregional disease (cases 3–5, 7, and 10; Table 1
in the Data Supplement), whereas five of the six children with widespread disease (cases 12, 14, 21, 22, and 24; Table 1
in the Data Supplement) had high concentrations (6000–7 930 000 transcripts/mL) in the bone marrow. The exception was patient 25 (stage 4M), who had undetectable tyrosine hydroxylase mRNA. The difference between the groups was statistically significant (P <0.05, Wilcoxon rank-sum test and two-sided Fisher exact test.).
first samples in the group already diagnosed and treated
Nine children were diagnosed and started on treatment before the first blood or bone marrow samples were collected. Three of them had locoregional disease, and six had widespread disease. In the former group, two children (cases 9 and 11; Table 1
in the Data Supplement) had no detectable tyrosine hydroxylase mRNA, and one child (case 1; Table 1
and Fig. 1
in the Data Supplement; available at http://www.clinchem.org/content/vol49/issue1/) had increased tyrosine hydroxylase mRNA in the blood and bone marrow.
Two of the six patients (cases 13 and 23; Table 1
in the Data Supplement) with widespread disease had increased tyrosine hydroxylase mRNA in the blood. Their samples were taken at relapse 18 and 28 months, respectively, after diagnosis. Four patients (cases 15–18; Table 1
in the Data Supplement) had undetectable blood tyrosine hydroxylase mRNA, but we obtained bone marrow samples from three of them that showed increased concentrations of tyrosine hydroxylase mRNA.
results during follow-up
Three of the 11 children with locoregional disease had increased blood concentrations of tyrosine hydroxylase mRNA. From one of the three, we did not obtain additional blood samples. The other two showed normalization of the mRNA concentrations during follow-up. One of these children also had a markedly increased tyrosine hydroxylase mRNA concentration in the bone marrow at diagnosis, which was also normalized during treatment. However, this patient (case 1) with unfavorable tumor biology had a subsequent metastatic relapse and died from progressive disease (Fig. 1
in the Data Supplement).
Nine children with widespread disease had increased blood concentrations of tyrosine hydroxylase mRNA, and in six of them the concentrations normalized during treatment. In one child (case 13; Fig. 2
in the Data Supplement; available at http://www.clinchem.org/content/vol49/issue1/) the concentration increased, whereas in two cases only one blood sample was obtained. Nine of the children with widespread disease had increased tyrosine hydroxylase mRNA concentrations in the bone marrow. In all patients followed with more than one blood sample, tyrosine hydroxylase mRNA decreased during treatment. In one case, the tyrosine hydroxylase mRNA was normalized after 30 months but had increased again 3 months later.
tyrosine hydroxylase MRNA in blood before, during, and after surgical manipulation of tumor
We collected serial blood samples from 10 children during surgery (Table 8
). Three children (cases 4, 7, and 19) had not received any preoperative treatment. Two patients had increased blood concentrations before surgery. Five of the 10 patients had increased concentrations at any time point before, during, or after surgery. In these children, the highest concentrations occurred during or 1 h after surgery, indicating the release of neuroblastoma cells to the circulation during surgical manipulation of the tumor (P <0.05, sign test).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsDiscussionReferences |
|---|
50 transcripts/mL of blood. At that low concentration, the imprecision of the PCR step is not possible to calculate because of the stochastic nature of the data (14); either the mRNA signal is there or not. We therefore calculated the detection limit from samples with 50–100 transcripts/mL of blood and found a mean SD of 36 transcript/mL. Thus the detectable concentration would be 72 transcripts/mL if 2 SD were used to calculate the detection limit.
Imprecision data from quantitative analysis of DNA and RNA are scanty in the literature, and in the case of real-time methods, CVs have been calculated from the mean and SD of the Cts at different concentrations of template (11). However, because these data regard the number of cycles needed to reach a certain threshold, they are not easy to interpret in terms of variation of measured transcripts. The data should be given in relation to the concentration of the number of target molecules. Using a calibrated competitive RT-PCR method for determination of MART-1 mRNA in blood, Sørensen et al. (15) found an intraassay imprecision of 15% and an interassay imprecision of 33% for 30 x10-21 mol of transcripts/mL (corresponding to 18 000 transcripts/mL). The authors demonstrated a detection limit of 1.5 x 10-21 mol/L (corresponding to 900 transcripts/mL). These data should be compared with the imprecision at different concentrations and the detection limit of 70 transcripts presented in this study.
The imprecision data should also be regarded in relation to the need for measuring differences between patient groups or between individual patients and to follow certain patients over time. From the patient data presented here, it is obvious that great variations are to be found within and between patients. Thus the results varied in our patients from <72 to 4.6 x 104 transcripts/mL in blood (Table 7A
), and from <72 to 8 x 106 transcripts/mL in bone marrow (Table 7B
). These data (see also Figs. 1
and 2
in the Data Supplement) illustrate that the imprecision does not always need to be very low. However, when for measuring smaller differences (Table 8
), low imprecision is needed.
The specificity of the method was evaluated by analysis of samples of cord blood and venous blood from healthy individuals. In none of these cases were results above the calculated detection limit. This indicates a high specificity of the method for blood analysis. With bone marrow we found that 5 of 19 children with leukemia had (moderately) increased concentrations. This indicates that bone marrow from leukemia patients is not a suitable matrix to evaluate the analytical specificity of the method. Bone marrow samples harvested from patients with solid tumors for autologous transplantation would be more suitable in this context (16). Usually there are no difficulties in differentiating between neuroblastoma and leukemia in children. However, the results are of interest because it has been shown that catecholamines are synthesized by lymphocytes (17) and that tyrosine hydroxylase immunoreactivity has been found in the cytoplasm and plasma membrane of lymphocytes (18).
The number of transcripts per tumor cell may vary from child to child; therefore, quantitative analysis of tyrosine hydroxylase is not an absolute measure of the number of neuroblastoma cells in different patients. However, we assume that the number of transcripts is reasonably constant in the tumor cell population of a patient. We therefore believe that the present method is appropriate for monitoring the progression and regression of micrometastases in children with neuroblastoma. However, tyrosine hydroxylase may be up-regulated in tumors treated with the angiogenesis inhibitor TNP-470 (19). The number of transcripts per cell therefore may change because of hypoxia in the tumor. This approach, therefore, should be studied more systematically with quantitative methods.
Several studies using qualitative methods have shown that tyrosine hydroxylase mRNA may be increased in children with neuroblastoma, both in blood and in bone marrow (7)(8)(20). In the present study, we show that children with widespread disease (stage 4) have significantly higher tyrosine hydroxylase mRNA in blood and bone marrow at diagnosis than those with locoregional disease (stages 1–3). This is in line with the fact that the prognosis in stages 1–3 is better than in stage 4 (2). In stage 4, children without demonstrable bone marrow involvement have a more favorable prognosis than those with infiltrated marrow (2).
The presence of neuroblastoma cells in the bone marrow detected by a sensitive immunocytologic analysis provides prognostic information (21). An increased tumor cell concentration in bone marrow at diagnosis as well as positive blood immunocytology at diagnosis is associated with decreased event-free survival (22). Similar results were obtained when bone marrow was analyzed with a quantitative method for GD2 synthase mRNA (23) and with a qualitative method for tyrosine hydroxylase mRNA (24). Furthermore, the presence of circulating neuroblastoma cells, as detected by analysis of tyrosine hydroxylase, is an independent poor prognostic indicator in children with stage 4 neuroblastoma (10). Thus, there are accumulating indications that MRD identified by sensitive methods is of great importance for identification of patients at high risk. The contribution from our study is that significant differences were observed between patients with stage 1–3 disease and those with stage 4 disease. The advantage of a quantitative measurement of tyrosine hydroxylase expression in terms of prognostic assessment must be tested in a larger cohort of patients with a longer follow-up.
Because the method we describe is quantitative, progression and regression could be followed in blood and bone marrow. Although our patient material was limited, there are several interesting cases to discuss. The results from two children are given in more detail in Fig. 1
and Fig. 2
in the Data Supplement (http://www.clinchem.org/content/vol49/issue1/). These cases illustrate how the analytical data closely followed the clinical course and that analysis of MRD may be of importance in presumed localized disease.
As shown by serial samples during surgery, the method described has the ability to detect minor changes in amounts of circulating transcript and the impact of reduction of tumor mass (Table 8
). The transcript concentration in blood after preoperative induction chemotherapy was normal in six of seven high-risk neuroblastoma patients (Table 8
, cases 9, 10, 16, 17, and 20–22). This indicates both that chemotherapy may eradicate systemic disease and that the described method is useful for monitoring this effect. Interestingly, the only child (case 22) among these seven who did not achieve a normalized preoperative transcript concentration in blood was the only one who relapsed and died from progressive disease (after conclusion of the present study). This indicates an adverse prognostic effect of remaining MRD after therapy, which is similar to results reported by Burchill et al. (10) using a qualitative RT-PCR method.
In conclusion, this quantitative tyrosine hydroxylase mRNA method is reliable and easy to perform. The children with stage 4 disease had very high amounts of transcript in the blood and bone marrow compared with children with stage 1, 2, or 3 disease. For some of the children, the transcript concentration was increased in the blood before the appearance of clinical symptoms or any other signs of relapse. It was also possible to monitor how the transcript concentrations diminished after treatment with chemotherapy, 131I-meta-iodobenzylguanidine, and surgery, respectively. We agree with Reynolds and Seeger (25) that detection of MRD may not only allow stratification of patients for clinical studies but also may provide a means for assessing the effect of novel therapies without the need to wait for the larger tumor burden necessary for quantification by routine clinical methods.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsDiscussionReferences |
|---|
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  J. Stutterheim, A. Gerritsen, L. Zappeij-Kannegieter, B. Yalcin, R. Dee, M. M. van Noesel, F. Berthold, R. Versteeg, H. N. Caron, C. E. van der Schoot, et al. Detecting Minimal Residual Disease in Neuroblastoma: The Superiority of a Panel of Real-Time Quantitative PCR Markers Clin. Chem., July 1, 2009; 55(7): 1316 - 1326. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Stutterheim, A. Gerritsen, L. Zappeij-Kannegieter, I. Kleijn, R. Dee, L. Hooft, M. M. van Noesel, M. Bierings, F. Berthold, R. Versteeg, et al. PHOX2B Is a Novel and Specific Marker for Minimal Residual Disease Testing in Neuroblastoma J. Clin. Oncol., November 20, 2008; 26(33): 5443 - 5449. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Y. Cheung, Y. Feng, A. Vickers, W. Gerald, and N.-K. V. Cheung Cyclin D1, a Novel Molecular Marker of Minimal Residual Disease, in Metastatic Neuroblastoma J. Mol. Diagn., April 1, 2007; 9(2): 237 - 241. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. V. Corrias, R. Haupt, B. Carlini, S. Parodi, L. Rivabella, A. Garaventa, V. Pistoia, and S. Dallorso Peripheral blood stem cell tumor cell contamination and survival of neuroblastoma patients. Clin. Cancer Res., October 1, 2006; 12(19): 5680 - 5685. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Berois, E. Blanc, H. Ripoche, X. Mergui, F. Trajtenberg, S. Cantais, M. Barrois, P. Dessen, B. Kagedal, J. Benard, et al. ppGalNAc-T13: A New Molecular Marker of Bone Marrow Involvement in Neuroblastoma Clin. Chem., September 1, 2006; 52(9): 1701 - 1712. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Mitsuhashi, S. Tomozawa, K. Endo, and A. Shinagawa Quantification of mRNA in Whole Blood by Assessing Recovery of RNA and Efficiency of cDNA Synthesis Clin. Chem., April 1, 2006; 52(4): 634 - 642. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Swerts, B. De Moerloose, C. Dhooge, J. Vandesompele, C. Hoyoux, K. Beiske, Y. Benoit, G. Laureys, and J. Philippe Potential Application of ELAVL4 Real-Time Quantitative Reverse Transcription-PCR for Detection of Disseminated Neuroblastoma Cells Clin. Chem., March 1, 2006; 52(3): 438 - 445. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S A Burchill Micrometastases in neuroblastoma: are they clinically important? J. Clin. Pathol., January 1, 2004; 57(1): 14 - 20. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
