.JPG?ad=18481&adview=true)
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Molecular Oncology, Markers, Clinical Correlates |
Centre for Cell and Molecular Medicine, School of Postgraduate Medicine, Keele University, Stoke on Trent, Staffordshire ST4 7QB, United Kingdom [E. K., M. V., D. S., J. B., W. E. F., R. N. C.]; Department of Endocrinology, Royal Victoria Hospital, Belfast, United Kingdom [A. B. A.]; Department of Endocrinology, Radcliffe Infirmary, Oxford, United Kingdom [J. A. H. W.]; Department of Endocrinology, University Medical Centre, Ljubljana, Slovenia [M. P.]; and Department of Endocrinology, St. James’s Hospital, Leeds, United Kingdom [P. B.]
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONPATIENTS AND METHODSRESULTSClonal Analysis of Recurrent...Relationship of LOH to...DISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSClonal Analysis of Recurrent...Relationship of LOH to...DISCUSSIONREFERENCES |
|---|
The initiating events in sporadic tumorigenesis of the pituitary are still largely unknown, although these adenomas are monoclonal (1, 2, 3) and therefore are thought to arise from de novo somatic mutations, providing a growth advantage to a single cell that, by clonal expansion, results in the adenoma. Apart from the somatotroph-specific Gsp oncogene, activation of classical oncogenes is not a common or early event in pituitary tumorigenesis (4, 5, 6) , occurring only rarely in pituitary carcinomas (6) . Other potential initiating mechanisms are loss of TSG3 function. Of the commonly studied TSGs, the retinoblastoma (Rb) gene is not deleted, despite LOH in this region of chromosome 13q in invasive tumors and pituitary carcinomas (7 , 8) . Similarly, the MEN-1 gene seems to be expressed in most sporadic tumors, despite frequent LOH of closely linked markers on 11q13 (9, 10, 11) . Also, LOH on 9p close to the MTS1/CDKN2A/p16 gene appears to be a frequent event (12) in about 30% of noninvasive and invasive pituitary tumors, suggesting that this is an early event in pituitary tumorigenesis, and this gene is also retained.
In the light of the above findings, we have proposed previously that pituitary tumor invasiveness (defined radiologically), and possibly biological aggressiveness, may be a consequence of an increasing accumulation of loss of TSG function (4) , and that the hypothesis (13) of accumulating genetic changes in advancing tumorigenesis may apply to pituitary tumors. We tested this hypothesis in specimens from patients with tumor recurrence or regrowth by comparing the pattern of LOH between successive samples. Moreover, this approach also provides an opportunity to determine whether a recurrent/regrown tumor is derived from the same or a different clone from the original tumor.
|
PATIENTS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSClonal Analysis of Recurrent...Relationship of LOH to...DISCUSSIONREFERENCES |
|---|
.
The indications for second and subsequent operations were variable,
including regrowth of residual tumor requiring debulking before
radiotherapy, regrowth sufficient to cause new or progressive visual
impairment despite radiotherapy, regrowth causing significant headache,
worsening or drug-resistant hormone hypersecretion in prolactinomas,
acromegaly, or Cushing’s disease. In the majority of patients, the
regrowth/recurrent tumor was large (>1-cm diameter). The patients
ranged in age from 16 to 78 years at first operation (32 males; 17
females). The interval between first and second operation varied from 1
to 25 years (median, 3 years). All cases were sporadic cases with
normocalcemia and no family history of endocrinopathy. No patient was
from a known MEN-1 family. Clinical subtypes were: 32 nonfunctioning; 7
somatotrophinomas; 5 prolactinomas; 4 corticotrophinomas; and 1
thyrotroph adenoma. All were confirmed by immunohistochemistry.
Thirty-six patients received radiotherapy at some stage, and in 21
patients at least 1, occasionally more, the sample was available for
analysis after radiotherapy. The extent of tumor invasiveness was based
on radiological criteria, defined by computed tomography or magnetic
resonance imaging scans judged by the individual clinical center, as
described previously (8)
. Most initial and recurrent
tumors were invasive, meaning the tumor was present in one or both
cavernous sinuses or invading into the sphenoid sinus. Blood samples
for constitutive DNA extraction were obtained from all subjects after
informed consent was obtained. The study was approved by the North
Staffordshire District and other institutional ethical committees.
|
LOH Analysis
Loss at known or putative TSG loci was assessed by microsatellite
analysis of microdissected tumor and matched patient blood DNA. The
rationale for selection of loci for analysis was based on previous
studies of frequent loss and/or association with invasive pituitary
adenomas or pituitary carcinomas (7
, 8
, 12
, 14
, 15)
. The
microsatellite markers on chromosomes 9p, 10q, 11q, and 13q, their map
position, primer sequences, and the PCR conditions have been described
in recent studies (8
, 12
, 14)
. Primer sequences, obtained
from the Genome Database, were designed to yield products of <200 bp.
PCRs were carried out in 25 µl volumes with 1.5 mM
MgCl2, 200 µM each of dATP, dGTP,
dTTP, and dCTP, 2 pmol of each primer, template DNA (as serial
dilutions), and 1 unit of Taq DNA polymerase. PCR was carried out for
25–28 cycles. Constitutive and tumor DNA products were run adjacently
and separated on 8% nondenaturing polyacrylamide gels, fixed in 10%
methylated spirit/0.5% acetic acid for 6 min, and then incubated in
0.1% aqueous silver nitrate for 15 min. After two brief washes in
distilled water, products were visualized by development in 1.5%
sodium hydroxide/0.1% formaldehyde. All samples from the same patient
were PCR amplified in the same experiment and run in adjacent lanes on
the same gel.
Allelic loss was identified by a reduction in band intensity of >80% by visual inspection or the absence of one of the expected allele PCR products in the amplified DNA. All LOH assignments were agreed by two observers and confirmed by an independent observer unaware of the case sample order. In case of disagreement, the analysis was repeated, and if there was still doubt about assignment, the result was declared noninformative. In this way, the effect on the conclusion regarding assignment would be "neutral." We have shown previously in "mixing" experiments that 30% contamination with normal DNA is required to obscure allele loss (12) . Template DNA was serially diluted prior to PCR amplification, allowing a direct comparison of dilutions that produced similar product intensities between constitutive DNA and the retained allele(s) in the matched tumor sample. For samples showing LOH, this was confirmed by at least one, often two, repeat analyses on a separate DNA extract from additional slides from the same paraffin block. Where the block contained spatially discrete tumor fragments large enough, these were extracted separately to compare LOH within the same tumor specimen. We confirmed, in five such cases with at least one LOH, that this was identical in two or more fragments, confirming that for that the sample the cells within it were derived from the same clone. There were no samples with sufficient adjacent normal pituitary tissue to confirm that the LOH was confined to the tumor, although we have previously failed to show any LOH in normal pituitary tissue processed in an identical manner (8 , 12 , 14) .
Because we were concerned that extraction of DNA from paraffin-embedded
tissue might lead to preferential loss of the larger allele at a given
locus, we coamplified the housekeeping gene GAPDH from
several of the samples. The GAPDH amplicon was designed to
yield a product of 252 bp, which was larger than those amplified by the
microsatellite analysis. In all instances, the larger GAPDH
fragment amplified irrespective of which allele was lost. An example of
this is shown in Fig. 1
.
|
90%), short tandem
repeat within HUMARA (16)
. Tumor and matched
blood DNA were digested in a total volume of 10 µl for 16 h with
10 units of the methylation-sensitive restriction enzyme
HpaII. To ensure complete digestion of the DNA, an
additional incubation (16 h) with an additional 10 units of
HpaII restriction enzyme was performed. Approximately 200 ng
of digested and undigested DNA were subjected to a primary PCR
amplification of a region that encompasses a HpaII
methylation restriction site in addition to the HUMARA
trinucleotide ([CAG]n) repeat element. PCR amplification was achieved
using specific oligonucleotides (sense, 5'-
GCTGTGAAGGTTGCTGTTCCTCAT-3'; antisense, 5'-TCCAGAATCTGTTCCAGAGCGTGC-3')
and limited to 12 cycles. One-tenth volume of this PCR reaction was
then used in a secondary hemi-nested PCR amplification and limited to
25 cycles (sense, 5'- GCGTGCGCGAAGTGATCC-3'). PCR products were
electrophoresed on 8% nondenaturing polyacrylamide gels and
visualized as described above.
PGK-1 Gene.
In addition, clonality of initial and subsequent tumor samples was
analyzed using the PGK-1 PCR-based method of clonal
analysis. This method is based on a RFLP of the X chromosome
phosphoglycerokinase (PGK-1) gene and on the differential
methylation of the PGK-1 gene because of random inactivation
of one of two X chromosomes in females heterozygous for the
BstX1 polymorphism (17)
.
Approximately 200 ng of undigested or HpaII-digested DNA samples from female patients (see above) were used for PCR amplification of the PGK-1 locus with primers 1A (sense, 5'-CTGTTCCTGCCCGCGCGGTGTTCCGCATTC-3') and 1B(antisense, 5'-ACGCCTGTTACGTAAGCTCTGCAGGCCTCC-3'). Samples were amplified for 12 cycles, and one-tenth volume of this reaction was used in a secondary PCR reaction with internal primers 2A (sense, 5'-AGCTGGACGTTAAAGGGAAGCGGGTTCGTTA-3') and 2B (antisense, 5'-TACTCCTGAAGTTAAATCAACATCCTCTTG-3') and amplified for 25 cycles. After amplification, samples were digested with 10 units of BstXI restriction enzyme for 16 h. Restriction fragments were electrophoresed on 6% nondenaturing polyacrylamide gels and visualized by silver staining. Patients were deemed informative for the PGK-1 polymorphism if PCR amplifications and subsequent digestion demonstrated the presence of two fragments of 433 and 530 bp, resulting from alleles with and without the BstXI polymorphism, respectively. However, when samples were first digested with HpaII prior to amplification and subsequent post-PCR digestion with BstXI, a single band indicative of monoclonality was observed. Assignment of X chromosome allelic loss was on the same basis as autosomal LOH.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSClonal Analysis of Recurrent...Relationship of LOH to...DISCUSSIONREFERENCES |
|---|
and 4
,
and representative examples of LOH are shown in Fig. 2
. Figs. 3
and 4
shows the pattern of allelic loss
in the first and subsequent tumors for each of the 33 patients
exhibiting LOH. On the basis of LOH pattern in the first and subsequent
tumors, two common patterns were seen: pattern A, where LOH observed in
the original tumor was identical (preserved) in the patients subsequent
tumor(s) (Fig. 2
, T16), and in some cases this was accompanied by
additional losses (Fig. 3
, e.g., T7, T9, T12, and T23); and
pattern B, where LOH was observed in the first sample, and both alleles
were retained (loss to retention) in a subsequent sample(s) (Fig. 2
,
e.g., T 27 and T43). In some cases, this was accompanied by
additional losses (Fig. 4)
. Figs. 3
and 4
show that 14 patients showed
loss pattern A, and 19 patients showed loss pattern B (loss to
retention), respectively.
|
|
|
and 3)
. For example, the three
samples from subject 16 are likely to be from the same clone because of
loss of marker D13S1246 on all occasions and retention of
all other informative markers. Equally, pattern A is seen in subject
23, where loss at the marker D13S153 is seen in the first
(23.1) and subsequent (23.2) tumor, with an additional loss at the
marker PYGM, consistent with a progressive accumulation of
losses with time (Fig. 3)
. In some cases, failed determinations of LOH
confounded this interpretation. For example, in patients 28, 34, and
40, LOH analysis (Fig. 3)
at markers that failed in the first tumor
revealed loss in the subsequent tumor. We therefore assigned these
patients to loss pattern A because we consider that these would be
similar to patients 13 and 15 if the marker was retained, or to
patients 16, 21, and 32 if the marker was lost. The clinical
characteristics of the patients were 10 males (ages, 20–57) and 4
females (ages, 25–34). Tumor subtype distribution was 8 nonfunctioning
tumors (nos. 3, 7, 9, 16, 21, 23, 28, and 32); 3 somatotrophinomas
(nos. 15, 34, and 44); 1 prolactinoma (no. 40); 1 corticotrophinoma
(no. 12); and 1 thyrotrophinoma (no. 13). The median interval between
first and second/third surgery was 5 years (range, 1–15 years).
Loss Pattern B: Evidence for a Multiclonal Origin.
In samples showing retention of heterozygosity in the second or
subsequent sample, this could most likely be explained by the second
sample being derived from a distinctly separate clone (Figs. 2
and 4)
.
For example in patient 6 (Fig. 4)
, the first tumor showed loss at 4 of
the microsatellite markers, whereas in the second sample (6.2),
retention of heterozygosity at three of these markers was observed,
making it highly unlikely that this sample is clonally related to the
first. The possibility that retention of heterozygosity is attributable
to contamination by normal tissue is excluded because LOH is still
found in tumor 6.2 at the marker D13S1246. Equally, in
patient 1 (Fig. 4)
, the first tumor showed loss at three markers
(IFNA, D10S217, and PYGM), whereas all subsequent
tumors showed retention of heterozygosity at one of these markers
(IFNA). However, in this case, the phenomenon of "loss to
retention" is repeated between the third (1.3) and fourth (1.4)
operation, with retention of heterozygosity at the marker
PYGM. Thus, the allelic loss pattern in this case is
suggestive of the presence of three independent clones. For two tumors,
we were able to obtain discrete separate biopsies from the same
operation, and the demonstration of identical loss and retention
pattern across the nine markers argues against different loss patterns
existing in the same tumor and further supports the robustness of the
technique (data not shown).
The clinical characteristics of patients with this loss pattern were 12 males (ages, 17–78) and 7 females (ages, 16–45). Tumor subtype distribution was 12 nonfunctioning tumors (nos. 4, 5, 6, 17, 19, 22, 24, 26, 27, 31, 33, and 43); 2 somatotrophinomas (nos. 30 and 49); 3 prolactinomas (nos. 14, 35, and 36); and 2 corticotrophinomas (nos. 1 and 2). The median time between first and second/subsequent surgery was 3 years (range, 1–25 years).
|
Clonal Analysis of Recurrent Pituitary Tumors Assessed by X Chromosome Inactivation |
|---|
|
TopABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSClonal Analysis of Recurrent...Relationship of LOH to...DISCUSSIONREFERENCES |
|---|
. The
HUMARA or PGK-1 polymorphism confirmed a
different clonal origin in successive tumors in 5 of 11 patients (Fig. 5)
, examples of which are shown in Fig. 2
(T27 and T43).
HUMARA revealed two independent clones in patient 18,
evident as the lower allele in the first tumor and the higher allele in
the subsequent sample. However, in this case, no losses were found
across the autosomal microsatellite markers used here.
|
Two further tumors (nos. 34 and 43) were informative with
PGK-1. Tumor 34 showed the presence of an independent clone
in the second sample; however, in this case, LOH analysis was unable to
unambiguously define clonality because INFA failed to amplify in 34.1.
In tumor 43, PGK showed the presence of an independent clone
in the second sample (43.2; Fig. 2
). The presence of independent clones
was once again entirely consistent with the LOH studies that showed a
loss to retention pattern (pattern B).
Four additional subjects were informative for the HUMARA
polymorphism; three showed the same clone in each case in successive
tumors (nos. 20, 40, and 49). In two cases (nos. 20 and 40), the LOH
and X chromosome inactivation interpretation of single monoclonal
origin were entirely consistent. In subject 49, the LOH data (pattern
B) is consistent with two independent clones, but X chromosome
inactivation failed to confirm this. The likely explanation is that
49.1 and 49.2 do represent different clones, but in this case the same
X chromosome is inactivated. In patients 2 and 5, the first tumors were
monoclonal, as evidenced by a single lower band (Fig. 5
, band
L), whereas the subsequent tumors (2.2 and 5.2) gave a pattern
consistent with the presence of more than one clone (Fig. 5
,
bands L and H). The most likely interpretation of
these data is the presence of two clones in the second samples, one of
which has not sustained LOH at the microsatellite marker
D13S1246. Thus, the presence of this coexisting clone would
mask and negate the loss seen in the first tumors. In subject 45, the
initial and subsequent tumor is either biclonal or perhaps multiclonal.
The simplest explanation, in this case, is the coexistence of at least
two clones in each tumor sample not sustaining loss at any of the
markers used in this study. Conversely, the individual clones have
sustained loss at some of these markers, but these losses are not
coincident. In this case, retention in one clone and loss in the other
would confound microsatellite analysis in the determination of LOH. In
summary, X inactivation analysis revealed 5 of 11 patients with tumors
of different clonal origins, 3 of 11 from the same clone, and 3 of 11
were composed of two clones.
|
Relationship of LOH to External Radiotherapy |
|---|
|
TopABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSClonal Analysis of Recurrent...Relationship of LOH to...DISCUSSIONREFERENCES |
|---|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSClonal Analysis of Recurrent...Relationship of LOH to...DISCUSSIONREFERENCES |
|---|
In this study of recurrent pituitary tumors, we anticipated that recurrent tumors would be derived from the same clone as the original tumor with perhaps additional LOH, as predicted from the Fearnon and Vogelstein (13) model of tumorigenesis, where increasing LOH is associated with increased propensity to progression/recurrence. However, not all tumors conformed to this pattern because in some tumors, we observed a pattern of loss to retention between first and subsequent tumors. One possible explanation for the observed LOH patterns was that a proportion of recurrent tumors were derived from entirely independent clones. This was confirmed by the X chromosome inactivation analysis in women. There are at least two interpretations of these findings:
(a) Human pituitary tumors are multiclonal from the outset. Initially one clone dominates and the second clone cannot be identified. In this interpretation, the initiating event(s) might give rise to a population of predisposed cells that require a second "hit or insult." The second "hit or insult" gives rise to several clones at a single point in time. Further genetic abnormalities give one clone a growth advantage over another, leading to a dominant clone. When the tumor is removed at first surgery, cells from the dominant clone are mostly eliminated. However, if there is a remnant after surgery, this may comprise an admixture of cells from the original dominant clone and another clone, either of which are capable of expansion. Accordingly, cells in a tumor regrowth or recurrence may be predominantly from the original clone or another independent clone that was not initially dominant. In the latter case, the recurrent tumor is still monoclonal but from a different clone. However, if more than one clone (the original and second emerging clone) now expand equally, LOH and X chromosome inactivation will define the tumor as biclonal or polyclonal. If pituitary tumors do contain several independent clones ab initio, it may be possible to demonstrate this by showing different LOH and X inactivation patterns in multiple biopsies from the same tumor. However, in this retrospective study, this was only possible in two subjects, and both samples showed an identical LOH pattern.
(b) Recurrent tumors may develop entirely independently at a different point in time, and in this case different clones are temporally separated. Thus, while in the first scenario the second "insult" might target multiple cells at the same point in time, in the second, this event would be temporally discrete, targeting a single cell. In both cases, the second event is responsible for the growth advantage, allowing clonal expansion into a monoclonal tumor. From our data, it is not possible to ascertain which of these two scenarios applies, but that either is possible is shown for the first time from this analysis. In fact, we found that one of the aforementioned explanations was more likely than the progressive accumulation of additional allelic loss within the same original clone, because loss pattern B was more common than loss pattern A.
Thus, we suggest that recurrent human pituitary tumors arise from a separate independent clone in a significant proportion of cases (up to 60%), although within the same operative specimen they are genetically homogeneous, as evidenced by the identical LOH/X chromosome inactivation pattern in each individual aliquot. Furthermore, this conclusion applies to tumors derived from all of the cell lineages within the pituitary, indicating that whichever initiating scenario applies, this is likely to be similar for all tumor subtypes. It remains unclear why, in the same subject, an initially minor clone could become dominant. One possibility could be that the initially dominant clone produces inhibitory growth factors (19) which, in a paracrine fashion, prevent the outgrowth of the minor clone. When these paracrine inhibitory factors are removed by surgery, the minor clone may then be capable of expansion.
It is, at present, unclear what causes the earliest changes that constitute initiating events in pituitary tumorigenesis. Oncogenic mutations are not common, being restricted to a subset of somatotrophinomas only (Gsp; reviewed in Refs. 4 and 6 ). The pituitary tumor transforming gene (PTTG; Ref. 20 ), identified from a clonal rat pituitary cell line, may also play a significant role in human pituitary tumorigenesis and shows increased expression in invasive tumors (21) . Although LOH has been identified in the region of recognized tumor suppressor genes, MEN-1 and Rb, neither of these genes are deleted or mutated in the vast majority of sporadic tumors (7, 8, 9, 10) . Thus, LOH on chromosomes 11q13 and 13q point to as yet unidentified TSGs in these regions, although their overall frequency of loss (30–35%) makes it unlikely that these represent the earliest genetic changes responsible for tumor initiation.
To our knowledge, this is the first report that describes the presence of an emerging independent clone(s) in benign pituitary tumors from an individual subject as assessed by LOH and confirmed by X chromosome inactivation analysis. In three subjects, X chromosome inactivation showed a biclonal pattern that may reflect the presence of at least two codominant clones, each of which has a different X chromosome inactivated. However, these were the only cases in which we could not definitively exclude the possibility that this was attributable to contaminating normal tissue, but our use of microdissected, histologically defined tumor tissue makes this unlikely. In a previous report examining the clonal composition of corticotrophinomas, which were histologically discrete adenomas with no contaminating normal tissue, three of nine showed a polyclonal pattern, indicating that a proportion of pituitary adenomas may be biclonal/polyclonal (22) .
It might be expected that the DNA-damaging effects of radiotherapy could have resulted in some of the allelic losses observed and possibly the emergence of new clones. Although this might be feasible in some cases, our results show that this is by no means universal because nearly 60% of patients in whom we propose an independent clone received either no radiotherapy, or the treatment after the sample revealing the separate clone was identified.
There was no difference between the clinical characteristics of patients whose recurrent/regrown tumors appeared to be from an independent clone and those whose were from the same clone, especially with respect to age, sex, tumor subtype distribution, and time between operations. Thus, the observed differences in clonal origins were independent of these factors.
|
|
FOOTNOTES |
|---|
1 We thank the North Staffordshire National Health
Service Trust and the West Midlands Regional Research and Development
Directorate for financial support and provision of facilities. 
2 To whom requests for reprints should be
addressed, at Centre for Cell and Molecular Medicine, School of
Postgraduate Medicine, Keele University, Thornburrow Drive, Hartshill,
Stoke-on-Trent, Staffordshire ST4 7QB, United Kingdom.
3 The abbreviations used are: TSG, tumor
suppressor gene; LOH, loss of heterozygosity; MEN-1, multiple endocrine
neoplasia type 1; GAPDH, glyceraldehyde-3-phosphate dehydrogenase;
HUMARA, human androgen receptor allele.
Received 6/ 8/00; revised 8/ 7/00; accepted 8/ 8/00.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSClonal Analysis of Recurrent...Relationship of LOH to...DISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  E. V. Dimaraki, W. F. Chandler, M. B. Brown, C. A. Jaffe, S. Y. Kim, R. Taussig, V. Padmanabhan, and A. L. Barkan The Role of Endogenous Growth Hormone-Releasing Hormone in Acromegaly J. Clin. Endocrinol. Metab., June 1, 2006; 91(6): 2185 - 2190. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. P. Heaney Pituitary tumour pathogenesis. Br. Med. Bull., January 1, 2006; 75-76: 81 - 97. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. J. Simpson, A. M. McNicol, D. C. Murray, A. Bahar, H. E. Turner, J. A. H. Wass, M. M. Esiri, R. N. Clayton, and W. E. Farrell Molecular Pathology Shows p16 Methylation in Nonadenomatous Pituitaries from Patients with Cushing's Disease Clin. Cancer Res., March 1, 2004; 10(5): 1780 - 1788. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Mora, N.-K. V. Cheung, S. Oplanich, L. Chen, and W. L. Gerald Novel Regions of Allelic Imbalance Identified by Genome-wide Analysis of Neuroblastoma Cancer Res., March 1, 2002; 62(6): 1761 - 1767. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. J. Marx and L. K. Nieman Aggressive Pituitary Tumors in MEN1: Do They Refute the Two-Hit Model of Tumorigenesis? J. Clin. Endocrinol. Metab., February 1, 2002; 87(2): 453 - 456. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
, retentions. NI,
noninformative samples; F, failed to amplify. Tumor
subtype by immunohistochemistry: nonfunctional cases, 3, 7, 9, 16, 21,
23, 28, and 32; somatotrophinomas 15, 34, and 44; prolactinoma, 40;
corticotrophinoma, 12; thyrotrophinoma, 13. Males, 3, 9, 12, 13, 15,
16, 21, 28, 32, and 44; females, 7, 23, 34, and 40.
