Learn more: PMC Disclaimer | PMC Copyright Notice
Measurements of Plasma Methoxytyramine, Normetanephrine and Metanephrine as Discriminators of Different Hereditary forms of Pheochromocytoma
Abstract
BACKGROUND
Pheochromocytomas are rare catecholamine–producing tumors derived in at least 30% of cases from mutations in 9 tumor-susceptibility genes identified to date. Testing of multiple genes at considerable expense is often undertaken before a mutation is detected. This study assessed whether measurements of plasma metanephrine, normetanephrine and methoxytyramine, the O-methylated metabolites of catecholamines, might help distinguish different hereditary forms of the tumor.
METHODS
Plasma concentrations of O-methylated metabolites were measured by liquid chromatography with electrochemical detection in 173 patients with pheochromocytoma, including 38 with multiple endocrine neoplasia type 2 (MEN 2), 10 with neurofibromatosis type 1 (NF1), 66 with von Hippel-Lindau (VHL) syndrome and 59 with mutations of succinate dehydrogenase (SDH) type B or D genes.
RESULTS
In contrast to patients with VHL and SDH mutations, all patients with MEN 2 and NF1 presented with tumors characterized by increased plasma concentrations of metanephrine (indicating epinephrine production). VHL patients usually showed solitary increases in normetanephrine (indicating norepinephrine production), whereas additional or solitary increases in methoxytyramine (indicating dopamine production) characterized 70% of patients with SDH mutations. Patients with NF1 and MEN 2 could be discriminated from those with VHL and SDH mutations in 99% of cases by the combination of normetanephrine and metanephrine. Measurements of plasma methoxytyramine discriminated patients with SDH mutations from those with VHL mutations in a further 78% of cases.
CONCLUSIONS
The distinct patterns of plasma catecholamine O-methylated metabolites in patients with hereditary pheochromocytoma provide an easily utilized tool to guide cost-effective genotyping of underlying disease-causing mutations.
Introduction
Pheochromocytomas and paragangliomas (PPGLs) are rare catecholamine-producing tumors that arise respectively from adrenal and extra-adenal chromaffin tissue and which occur sporadically or as a result of germ-line mutations of several tumor susceptibility genes (1). Mutations of the rearranged during transfection (RET) gene in multiple endocrine neoplasia type 2 (MEN2), the von Hippel-Lindau (VHL) gene in VHL syndrome, the neurofibromatosis type 1 (NF1) gene in von Recklinghausen disease and of genes encoding succinate dehydrogenase (SDH) subunits B (SDHB) and D (SDHD) are the most well known causes of hereditary PPGLs. Mutations of the gene for SDH subunit C are a less frequent cause of catecholamine-producing PPGLs (2).
Reported frequencies of germline mutations of the above genes among patients with PPGLs range from 27% to 32% (3–5). This frequency is likely to increase as further tumor susceptibility genes are identified. Most recently mutations of genes encoding the SDH complex assembly factor 2, transmembrane protein 127 and SDH subunit A have been identified as further hereditary causes of PPGLs (6–8). This brings together a total of nine tumor susceptibility genes now recognized as responsible for hereditary PPGLs.
Findings that substantial proportions of germ-line mutations of tumor susceptibility genes occur in patients without a syndromic presentation or obvious hereditary basis for the tumors provide an argument for genetic testing of all patients with PPGLs (4, 9). It is nevertheless widely recommended that, although there is now a reasonable argument for more widespread genetic testing, it is neither appropriate nor currently cost-effective to test every disease-causing gene in all patients with PPGLs; rather the decision to test and which genes to test requires judicious consideration of numerous factors (10–12). The catecholamine biochemical phenotype of tumors was suggested as one factor potentially useful to guide decision-making (12), but available data to support this link remain limited.
Importantly, any attempt to use diagnostic data to examine catecholamine phenotypes of PPGLs should ideally utilize the most sensitive diagnostic tests available. Because these tumors do not always secrete catecholamines, measurements of these analytes in plasma or urine often fail to reveal the presence of the tumors (13). In contrast, PPGLs continuously metabolize catecholamines to the O-methylated metabolites by a process that is independent of variations in catecholamine release (14). Thus, measurements of plasma free normetanephrine and metanephrine, the respective O-methylated metabolites of norepinephrine and epinephrine, represent the most sensitive tests to diagnose catecholamine production by PPGLs (13). Additional measurements of methoxytyramine, the O-methylated metabolite of dopamine, provide a further useful analyte for indicating tumor production of dopamine (15).
The present analysis utilized a dataset from a large population of patients with hereditary PPGLs, linked to a tumor tissue bank, to examine whether differences in plasma concentrations of the O-methylated metabolites might be useful for distinguishing the five main hereditary forms of PPGLs. The aim was to establish whether such information derived from diagnostic testing might offer utility as a guide for genotyping patients with PPGLs.
Materials and Methods
PATIENTS
The study involved retrospective analysis of data from 173 patients with hereditary PPGLs. Most subjects were investigated at the National Institutes of Health (Bethesda, Maryland, US) and the others at European Centers, including Radboud University Medical Center (Nijmegen the Netherlands), the University of Florence (Florence, Italy) and Dresden University Hospital (Dresden, Germany). Written informed consent was obtained from patients enrolled into intramural review board approved studies at the NIH. At European centers, when informed consent was not obtained, the data were collected under conditions of regular clinical care, with ethical committee approval obtained for the use of those data.
Amongst the 173 patients with hereditary PPGLs, there were 66 with VHL syndrome, 38 with MEN 2, 10 with NF1, 48 with mutations of the SDHB gene and 11 with mutations of the SDHD gene. All patients with MEN 2 and NF1 and most with VHL syndrome were diagnosed with already established gene mutations or hereditary syndromes at the time of testing for catecholamine-producing tumors. However, one patient was identified with a VHL mutation, 37 with SDHB mutations and 8 with SDHD mutations as a consequence of routine testing of RET, VHL, SDHD and SDHB genes implemented after 2005.
COLLECTIONS OF BLOOD, URINE AND SURGICAL SPECIMENS
Blood samples from all patients were obtained with subjects supine for at least 20 min before blood collection. Subjects were instructed to fast and abstain from caffeinated and decaffeinated beverages overnight and avoid taking acetaminophen for 5 days before blood sampling. Samples of blood were transferred into tubes containing heparin as anticoagulant and immediately placed on ice until centrifuged (4°C) to separate the plasma. Plasma samples were stored at −80°C until assayed. Twenty-four hour urine samples were collected from 162 patients using HCL as a preservative. Total urine volume was determined and aliquots stored at 4°C until assayed.
Samples of tumor tissue were procured from 90 patients. Small samples of each tumor (10 to 50 mg) were dissected from the mass, frozen on dry ice and stored at −80°C. For further processing, tissue samples were weighed frozen and then homogenized in at least 5 volumes of 0.4 M perchloric acid containing 0.5 mM EDTA. Homogenized samples were centrifuged (1500 × g for 15 min at 4°C) and supernatants collected and stored at −80°C until assayed for catecholamines.
LABORATORY ANALYSES
Plasma concentrations of free metanephrine, normetanephrine and methoxytyramine were quantified by liquid chromatography with electrochemical detection as described previously (15, 16). Intra-assay coefficients of variation established from an inter-laboratory quality assurance program described by Pillai and Callen (17), varied over cycles of assessment between 2.2–7.1% for metanephrine, 5.9–8.7% for normetanephrine and from 8.0–23.9% for methoxytyramine.
Concentrations of catecholamines (epinephrine, norepinephrine and dopamine) in plasma and tumor tissue were also quantified by liquid chromatography with electrochemical detection (18, 19). Twenty-four hour urinary outputs of catecholamines (epinephrine, norepinephrine and dopamine) and deconjugated (free plus conjugated) fractionated metanephrines (metanephrine and normetanephrine) were measured at outside laboratories by HPLC or by liquid chromatography with tandem mass spectroscopy as described previously (20).
Reference intervals for plasma concentrations of free metanephrine (0.06–0.31 nmol/L), free normetanephrine (0.10–0.61 nmol/L) and free methoxytyramine (0.006–0.090 nmol/L), epinephrine (4–83 0.02–0.45 nmol/liter), norepinephrine (0.47–2.95 nmol/L) and dopamine (0.013–0.379 nmol/L) were established from combined groups of 175 normotensive and 110 hypertensive (n=110) volunteers, as detailed elsewhere (13, 15). Reference intervals for 24 hr urinary outputs of deconjugated metanephrine (0.22–1.32 μmol/24 hr), deconjugated normetanephrine (0.70–2.64 μmol/24 hr), free epinephrine (0–0.11 μmol/24 hr), free norepinephrine (0.09–0.47 μmol/24 hr) and free dopamine (0.39–2.63 μmol/24 hr) were those provided by the outside laboratories responsible for those measurements.
STATISTICS
Due to the skewed distributions of plasma and urinary biochemical measurements, these data underwent logarithmic transformation before statistical analyses. Differences between multiple groups were assessed by one-way ANOVA. Post-hoc tests were performed using the Tukey-Kramer test. Stepwise linear discriminant analysis was carried out to determine which combinations of biomarkers optimally classified patients according to their hereditary condition. Principal components analysis was used to select combinations of the three most appropriate biomarkers that clustered data in n-dimensional space separately according to hereditary condition.
Results
TUMOR TISSUE CATECHOLAMINES
Tumor tissue concentrations of catecholamines showed considerable variability among the different groups of patients with PPGLs (Fig. 1). The presence of markedly higher tumor tissue contents of epinephrine in patients with MEN 2 and NF1 than in those with mutations of VHL, SDHB and SDHD genes represented the clearest distinguishing feature. In patients with VHL, SDHB and SDHD gene mutations, tumor tissue contents of epinephrine averaged less than 2% of the total combined contents of all catecholamines, whereas in patients with MEN 2 and NF1, epinephrine represented on average 50% and 43% of total tumor catecholamine contents respectively. Tumor tissue concentrations of dopamine were generally a minor component of tissue catecholamines, representing less than 1% of the total contents of catecholamines in all groups of patients except those with SDHB mutations. Tissue dopamine concentrations in tumors from these patients represented on average 26% of the total contents of all catecholamines, but showed considerable variation, ranging from less than 1% to 95% of the total.
Bar graphs showing tumor tissue contents of norepinephrine (panel A), epinephrine (panel B) and dopamine (panel C) in patients with MEN 2, NF1 and mutations of VHL, SDHB and SDHD genes. Results are expressed as mean (±SEM) percent values of total contents of all three catecholamines. Differences (P<0.05) between groups are indicated by different symbols (* and †), where * indicates significant higher values than † (Tukey-Kramer post-hoc test)
PLASMA AND URINARY NORMETANEPHRINE AND NOREPINEPHRINE
Considered alone, plasma concentrations of normetanephrine and norepinephrine showed no obvious distinguishing differences among the five groups of patients with hereditary PPGLs (Fig. 2A&B). Nevertheless, measurements of plasma normetanephrine were distinctly superior to those of norepinephrine for indicating tumoral norepinephrine production. More specifically, plasma concentrations of norepinephrine were below the upper limits of the reference population in 54 of the 173 patients with hereditary PPGLs (31%) compared to only 17 patients (10%) for plasma normetanephrine. Similarly urinary outputs of norepinephrine were normal in 34% of patients compared to only 14% of patients with normal urinary outputs of normetanephrine (Table 1).
Dot plots showing normalized logarithmic distributions of plasma concentrations of catecholamines and the free O-methylated catecholamine metabolites in patients with PPGLs due to VHL syndrome, MEN 2, NF1 and mutations of SDHB and SDHD genes. Plasma concentrations of normetanephrine (panel A) and its catecholamine precursor, norepinephrine (panel B), are shown in the upper panels; those of metanephrine (panel C) and its catecholamine precursor, epinephrine (panel D), in the middle panels; while concentrations of methoxytyramine (panel E) and its catecholamine precursor, dopamine (panel F), are shown in the lower panels. The dashed horizontal lines show the upper limits of reference intervals for each analyte.
Table 1
Proportions of Patients with Hereditary PPGLs and Elevated Urinary Outputs of Metanephrines or Catecholamines*
| Urinary Metanephrines | Urinary Catecholamines | ||||
|---|---|---|---|---|---|
| Normetanephrine | Metanephrine | Norepinephrine | Epinephrine | Dopamine | |
| VHL | 91% (32/35) | 11% (4/35) | 65% (42/65) | 2% (1/65) | 6% (4/62) |
| MEN 2 | 92% (22/24) | 96% (23/24) | 47% (16/34) | 59% (20/34) | 24% (6/25) |
| NF1 | 100% (7/7) | 100% (7/7) | 88% (7/8) | 75% (6/8) | 33% (2/6) |
| SDHB | 76% (34/45) | 16% (7/44) | 74% (32/43) | 2% (1/42) | 38% (16/42) |
| SDHD | 88% (7/8) | 0% (0/8) | 86% (6/7) | 0% (0/7) | 50% (3/6) |
| ALL | 86% (102/119) | 35% (41/118) | 66% (103/157) | 18% (28/156) | 22% (31/141) |
PLASMA AND URINARY METANEPHRINE AND EPINEPHRINE
In contrast to the general lack of distinguishing differences for plasma normetanephrine among the five patient groups (Fig. 2A), plasma concentrations of metanephrine almost completely distinguished patients with MEN 2 and NF1 from those with mutations of VHL, SDHD and SDHB genes (Fig. 2C). In particular, whereas patients with MEN 2 and NF1 showed pronounced increases in plasma concentrations of metanephrine, patients with VHL, SDHD and SDHB mutations showed mainly normal plasma concentrations of metanephrine. Nevertheless, 12 out of the 125 (10%) patients with the latter mutations had increases in plasma metanephrine above the upper reference limits resulting in some overlap with concentrations in MEN 2 and NF1 patients.
Plasma concentrations of metanephrine were clearly superior to epinephrine (Fig. 2C&D) for indicating tumoral epinephrine production and distinguishing patient groups according to underlying mutations. Of note, while plasma concentrations of metanephrine were elevated in all (100%) patients with MEN 2 or NF1, plasma concentrations of epinephrine were increased above reference intervals in only 24 (51%) of these patients.
Urinary outputs of metanephrine were similarly more effective than outputs of epinephrine for indicating the presence of epinephrine-producing tumors in NF1 and MEN 2 patients (Table 1); 97% of patients in these two groups had elevations of urinary metanephrine compared to 62% with elevations of epinephrine.
PLASMA METHOXYTYRAMINE AND DOPAMINE
Plasma concentrations of methoxytyramine and dopamine showed additional patterns useful for distinguishing patients with SDH mutations from those with other hereditary syndromes (Fig. 2E&F). More specifically, 72% of patients with SDHB mutations and 67% with SDHD mutations had elevations of plasma free methoxytyramine, compared to 17% for patients with VHL mutations and 39% for patients with NF1 or MEN 2. Similarly, 42% of patients with mutations of SDHD or SDHB genes had elevations of plasma dopamine compared to 5% of patients with VHL syndrome, MEN 2 and NF1. Urinary outputs of dopamine were also elevated in 40% of patients with mutations of SDHD and SDHB genes, compared to 13% of patients with other gene mutations (Table 1).
BIOMARKER COMBINATIONS
The 10% of patients with VHL, SDHB or SDHD mutations who exhibited slight elevations of metanephrine above the upper reference intervals (Fig. 2A) also mainly had large increases in plasma normetanephrine so that proportional increases in metanephrine were minimal. Combinations of biomarkers were therefore considered for distinguishing the various groups of patients.
Discriminant analysis indicated that measurements of plasma metanephrine considered alone could be used to correctly classify 97% of patients into two groups: one group with MEN 2 and NF1 and the other with mutations of VHL and SDH genes (Table 2). Measurements of normetanephrine and methoxytyramine used alone or in combination offered no discriminant value for this classification. However, after either of both of these measurements were combined with measurements of plasma metanephrine there was further improvement with over 99% correctly classified. Measurements of plasma and urinary catecholamines were less effective than plasma metanephrines for correctly classifying patients into the two groups, whereas measurements of urinary metanephrine exhibited similar effectiveness to measurements of plasma metanephrine.
Table 2
Discriminant Analysis for Classification of Patients According to Neurochemical Profile
| Test or test combination | MEN 2 & NF1 vs. VHL & SDH | VHL versus vs. SDH |
|---|---|---|
| Percent Correctly Classified
| ||
| Plasma O-methylated metabolites | ||
| NMN | 47% | 60% |
| MN | 97% | 50% |
| MTY | 53% | 78% |
| MN & NMN | 99% | 59% |
| MN & MTY | 99% | 79% |
| NMN & MTY | 54% | 78% |
| NMN & MN & MTY | 100% | 78% |
| Plasma catecholamines | ||
| NE | 55% | 59% |
| EPI | 81% | 46% |
| DA | 47% | 61% |
| EPI & NE | 84% | 60% |
| EPI & DA | 82% | 59% |
| NE & DA | 53% | 65% |
| NE & EPI & DA | 85% | 70% |
| Urine Metanephrines | ||
| NMN | 50% | 66% |
| MN | 98% | 57% |
| NMN & MN | 98% | 64% |
| Urine catecholamines | ||
| NE | 57% | 60% |
| EPI | 92% | 62% |
| DA | 61% | 59% |
| EPI & NE | 94% | 61% |
| EPI & DA | 93% | 62% |
| NE & DA | 67% | 62% |
| NE & EPI & DA | 94% | 69% |
Abbreviations: NMN, normetanephrine; MN, metanephrine; MTY, methoxytyramine; NE, norepinephrine; EPI, epinephrine; DA, dopamine.
Measurements of plasma methoxytyramine provided the best single biomarker for further distinguishing VHL patients from patients with mutations of SDHB and SDHD genes, correctly classifying 78% patients into either of these two groups (Table 2). Discriminant scores were negligibly increased by further combinations of biomarkers and remained higher than scores achieved for tests of plasma or urinary dopamine in combination with other biomarkers.
As illustrated by 3-dimensional plots of plasma concentrations of normetanephrine, metanephrine and methoxytyramine, data from patients with MEN 2 and NF1 clustered together within a group distinct from patients with mutations of VHL and SDH genes (Fig. 3). Data from patients with VHL and SDH mutations also showed two separate but overlapping clusters. For patients with MEN 2 there were strong positive relationships for plasma concentrations of normetanephrine versus metanephrine (r=0.640, P<0.001), normetanephrine versus methoxytyramine (r=0.618, P<0.001) and metanephrine versus methoxytyramine (r=0.551, P<0.001). In contrast, for patients with VHL or SDH mutations, plasma concentrations of metanephrine showed no relationships with either of the two other biomarkers; positive relationships were only observed for plasma concentrations of normetanephrine versus methoxytyramine (VHL, r=0.577, P<0.001; and SDH, r=0.318, P=0.017).
Principle components analysis scatter plot illustrating clustering of patients with PPGLs due to MEN 2 or NF1 (red dots) compared to patients with VHL (blue dots) and SDH (green dots) mutations according to relationships of plasma concentrations of methoxytyramine (x-axis), metanephrine (y-axis) and normetanephrine (z-axis). Values shown on axes are in units nmol/L.
Discussion
This study establishes that measurements of the O-methylated metabolites of catecholamines, in addition to providing useful diagnostic biomarkers for PPGLs, also enable discrimination of different hereditary forms of these tumors. Findings of distinct profiles for catecholamine O-methylated metabolites among different groups of patients with hereditary tumors are supported by additional measurements of catecholamines in tumor tissue and novel data are presented showing that plasma methoxytyramine provides a particularly sensitive biomarker for indicating tumoral dopamine production in patients with mutations of SDHB and SDHD genes.
While PPGLs in MEN 2 and VHL patients are established to show striking differences in tumor contents, metabolism and secretion of norepinephrine and epinephrine (21, 22), until now it has not been clear whether other hereditary PPGLs are characterized by similarly distinct differences. As shown here, tumors in patients with NF1 share a similar adrenergic phenotype to those in patients with MEN 2. Both groups develop tumors containing considerable epinephrine leading to consistently elevated plasma concentrations of metanephrine, but with increases in plasma and urinary epinephrine in less than 60% of cases.
In contrast to MEN 2 and NF1 groups, we now show here that tumors from patients with VHL, SDHD and SDHB mutations are characterized by low tissue levels of epinephrine with corresponding usually normal levels of metanephrine and epinephrine in plasma and urine. When plasma concentrations of metanephrine are increased, the increases are minor compared to the proportionally much larger increases of normetanephrine. Discriminant analysis revealed that by additional considerations of normetanephrine and methoxtyramine, the measurements of plasma metanephrine enable virtually all patients with mutations of VHL and SDH genes to be distinguished from those with MEN 2 and NF1.
There is some evidence that tumors in patients with SDHB mutations are characterized by mainly norepinephrine production, which in some patients includes additional increases in plasma dopamine (23). That evidence, however, did not include measurements of plasma methoxytyramine or tumor tissue dopamine. The present study reports for the first time that over two-thirds of patients with either SDHB or SDHD mutations have increases in plasma methoxyramine and that in the former patients this reflects the tumor contents of dopamine. Consistent with these observations, plasma concentrations and urinary outputs of dopamine were also more often increased in patients with PPGLs due to SDH mutations than in other groups. Discriminant analysis indicated, however, that measurements of plasma and urinary dopamine are less useful than measurements of plasma methoxytyramine for distinguishing patients with PPGLs due to SDH mutations from other groups.
Routine testing of tumor susceptibility genes is now often carried out in patients with PPGLs even when there is no evidence of a familial syndrome, a result of findings that up to 24% of such patients harbor unsuspected germ-line mutations of these genes (4, 9, 24). Such testing is extremely expensive. Considerable effort is therefore being expended in establishing guidelines and algorithms for cost-effective genotyping of patients with PPGLs (5, 10, 11, 25–28).
Early age of presentation of tumors provides a commonly used justification for genetic testing. Beyond this and in the absence of suggestive clinical stigmata, the locations of tumors and presence of malignancy provide important clues about which genes should be first tested (4, 5, 10, 26, 29). Extra-adrenal tumor location, particularly when multifocal, justifies testing of SDH genes. Presence of malignancy mandates testing of the SDHB gene. More recently, immunohistochemical analyses of SDHB expression in resected tumor specimens has been advocated as a useful method to stratify patients with and without SDH mutations, and thereby guide cost-effective genetic testing (30, 31).
The data here indicate another approach to streamline targeted testing of underlying germline mutations based on readily available and easily utilized results of inexpensive biochemical tests used during diagnosis of PPGLs. For patients in whom there is no characteristic clinical stigmata or family history to guide genetic testing, patterns of increases in plasma normetanephrine, metanephrine and methoxytyramine provide useful information to determine the most appropriate genes to test. Testing of the RET gene is only warranted for patients with tumors characterized by increases of plasma metanephrine. In patients with solitary increases of plasma normetanephrine, mutations of VHL, SDHD and SDHB genes remain possible. Among these patients, further stratification for genotyping is possible from additional measurements of plasma methoxytyramine. For those patients found to have tumors that produce increases of plasma methoxytramine, with or without additional increases of normetanephrine, testing of SDHB and SDHD genes is warranted. In patients with solitary increases of normetanephrine, additional testing of the VHL gene can be considered.
Combined with consideration of other factors — such as tumor location, metastatic involvement and SDHB tumor tissue immunohistochemistry studies — the data derived from diagnostic testing offer a useful approach to select genes for testing. Validation of such utility requires additional well-considered prospective studies directed at patients without a familial or syndromic presentation in whom factors already proposed for stratification of patients for gene testing may also be utilized. From this it should be possible to establish formal cut-offs for ages, catecholamine phenotypes and other indices beyond which testing for specific genes is not clinically indicated.
In addition to providing a cost-effective solution for gene testing, the results of this study have other immediate implications for patient care. In patients with confirmed mutations of SDHB or SDHD genes who must undergo periodic surveillance for PPGLs, our data indicate that the effectiveness of biochemical screening may be enhanced by additional measurements of plasma methoxytyramine and dopamine. This can be expected to be particularly important for detecting the occasional patients with tumors producing exclusively or near exclusively dopamine (15). Similarly during screening for PPGLs in patients with MEN 2 or NF1, interpretation of biochemical test results should be focussed on measurements of both metanephrine and normetanephrine. In contrast, in VHL syndrome the focus should be on normetanephrine.
In summary, the present study establishes distinct profiles of plasma concentrations of normetanephrine, metanephrine and methoxytyramine among different groups of patients with hereditary PPGLs. Measurements of these biomarkers provide information for predicting underlying mutations that should be useful for efficient and cost-effective genotyping. The distinct profiles are also important to consider during periodic testing for PPGLs in patients who are at risk for the tumors because of underlying germline mutations of tumor susceptibility genes.
Acknowledgments
Thanks are extended to Thanh-Truc Huynh, Nan Qin, Stephanie Fliedner, Kathryn King and Tamara Prodanov for technical help or assistance with collections of patient materials and data. This study was supported by the intramural programs of the National Institute of Child Health and Human Development and the Center for Cancer Research, National Cancer Institute, at the National Institutes of Health, Bethesda, Maryland, USA, the Deutsche Forschungsgesellschaft, the Center for Regenerative Therapies Dresden and the Dresden Tumor Center.