2004 Summer Meeting - Amsterdam - The Pathological Society of ...
2004 Summer Meeting - Amsterdam - The Pathological Society of ...
2004 Summer Meeting - Amsterdam - The Pathological Society of ...
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
169<br />
Congenital Cellular Mesoblastic Nephroma: Low Sensitivity<br />
for Detection <strong>of</strong> the ETV6-NTRK3 Fusion Transcript in<br />
Archival Paraffin-Embedded Material<br />
NJ Sebire 1 , S Gibson 1 , J Anderson 2 , M Malone 1 , AD<br />
Ramsay 1<br />
1 Department <strong>of</strong> Paediatric Pathology, Great Ormond Street Hospital,<br />
London, United Kingdom, 2 Department <strong>of</strong> Molecular Oncology, Great<br />
Ormond Street Hospital, London, United Kingdom<br />
Introduction: Congenital mesoblastic nephroma (CMN) is a common renal<br />
tumour <strong>of</strong> infancy in which the cellular variant has been reported to be<br />
associated with ETV6-NTRK3 fusion transcript also described in infantile<br />
fibrosarcoma (IFS). Recently, this fusion transcript has been detected from<br />
formalin fixed paraffin-embedded archival samples <strong>of</strong> both CMN and IFS with<br />
varying success. <strong>The</strong> aim <strong>of</strong> this study is to examine the use <strong>of</strong> RT-PCR for<br />
detection <strong>of</strong> the ETV6-NTRK3 fusion transcript in archival cases <strong>of</strong> formalinfixed,<br />
paraffin-embedded CMN, cellular, mixed and classic types.<br />
Methods. Cases <strong>of</strong> CMN in whom paraffin blocks were available were<br />
anonymised, H&E sections examined by two pathologists and classified as<br />
classic, cellular or mixed type. In each case, sections from a representative<br />
block were cut, RNA extracted and RT-PCR carried out according to a standard<br />
protocol to detect the ETV6-NTRK3 fusion transcript.<br />
Results. In all 13 cases <strong>of</strong> CMN, adequate detection <strong>of</strong> ß-actin housekeeper<br />
gene product was obtained confirming RNA viability <strong>of</strong> fragments up to 200bp<br />
in size. Only two cases demonstrated presence <strong>of</strong> the ETV6-NTRK3 fusion<br />
transcript including one <strong>of</strong> the four cases <strong>of</strong> CMN (25%), one <strong>of</strong> the four cases<br />
<strong>of</strong> mixed CMN (25%) but none <strong>of</strong> the five cases <strong>of</strong> classical CMN.<br />
Conclusion. In paraffin embedded archival material, the ETV6-NTRK3 fusion<br />
transcript is detectable in a minority <strong>of</strong> cases <strong>of</strong> cellular and mixed type CMN<br />
using rt-PCR. <strong>The</strong> finding <strong>of</strong> a positive result provides diagnostic confirmation<br />
but, as with many molecular techniques, a negative result does not exclude the<br />
diagnosis, particularly if fresh tissue is not available.<br />
170<br />
An Audit <strong>of</strong> the Clinical Usefulness <strong>of</strong> Examining Bone<br />
Marrow Trephine Biopsies at Multiple Levels with CD56<br />
Immunohistochemistry for Staging <strong>of</strong> Neuroblastic Tumours<br />
NJ Sebire , D Rampling , S Williams , M Malone , AD Ramsay<br />
Department <strong>of</strong> Paediatric Pathology, Great Ormond Street Hospital,<br />
London, United Kingdom<br />
Introduction: Detection <strong>of</strong> metastatic disease is important for staging and<br />
treatment modification in cases <strong>of</strong> paediatric neuroblastic tumour. In one case,<br />
neuroblastic tumour was not present on the initial trephine sections but a small<br />
focus became apparent in a subsequent level using immunohistochemistry.<br />
Following this case, our protocol has been to perform examination <strong>of</strong> trephine<br />
specimens for this indication at several levels with CD56<br />
immunohistochemistry on each level.<br />
Methods. A retrospective review <strong>of</strong> all cases <strong>of</strong> bone marrow trephine biopsies<br />
examined for the indication <strong>of</strong> neuroblastic tumour was carried out over a 17<br />
month period.<br />
Results. <strong>The</strong>re were 56 specimens examined for determination <strong>of</strong> involvement<br />
by neuroblastic tumour during the study period. In 12 cases (21%), neuroblastic<br />
tumour was present and in 44 cases (79%) no neuroblastic tumour was present.<br />
<strong>The</strong>re were no cases in which examination at multiple levels or use <strong>of</strong> routine<br />
immunohistochemistry aided either diagnosis in this series. If the index case is<br />
included, levels and immunostaining provide additional diagnostic information<br />
in a maximum <strong>of</strong> only 1.7% (95% CI 0%-9%) <strong>of</strong> such cases. <strong>The</strong> cost <strong>of</strong> levels<br />
and immunostaining (based on WELCAN units) is >200% greater than routine<br />
processing and staining, meaning it would cost at least £24,000 to possibly<br />
detect one extra case <strong>of</strong> marrow involvement. We therefore believe this is not a<br />
cost-effective protocol.<br />
Conclusion. <strong>The</strong> routine use <strong>of</strong> multiple levels and CD56<br />
immunohistochemistry in bone marrow trephine biopsies for assessment <strong>of</strong><br />
involvement by neuroblastic tumour is not cost-effective and only specifically<br />
indicated use <strong>of</strong> these techniques is recommended.<br />
171<br />
Immunohistochemical Findings in Embryonal ‘Small Round<br />
Cell’ Tumours with Molecular Diagnostic Confirmation<br />
NJ Sebire , S Gibson , D Rampling , S Williams , M Malone ,<br />
AD Ramsay<br />
Department <strong>of</strong> Paediatric Pathology, Great Ormond Street Hospital,<br />
London, United Kingdom<br />
Introduction: Diagnosis <strong>of</strong> pediatric tumors relies heavily on<br />
immunohistochemical staining <strong>of</strong> small tissue biopsies since many entities<br />
share a ‘small blue cell’ phenotype. Since introduction <strong>of</strong> molecular genetic<br />
analysis for detection <strong>of</strong> gene fusion products we have noted that tumors with<br />
proven molecular diagnoses can exhibit unusual patterns <strong>of</strong><br />
immunohistochemical staining.<br />
Methods. A panel <strong>of</strong> immunohistochemical stains was performed on sections<br />
from paediatric tumors with a ‘small blue cell’ phenotype in which molecular<br />
diagnostic confirmation was available. (S100, CD56, NB84, CD99 (MIC2),<br />
Bcl-2, CD117, CD34, Desmin, MNF116 and WT1).<br />
Results. In 370 sections from 37 cases, all PNETs, with and without the<br />
presence <strong>of</strong> t(11;22), demonstrated uniform membranous staining with CD99<br />
(MIC2) and focal staining with CD56, NB84, MNF116 and WT1. All<br />
rhabdomyosarcomas, both alveolar and embryonal, demonstrated uniform<br />
Desmin, CD56 and cytoplasmic WT1 immunostaining. DSRCTs showed<br />
positive cytokeratin staining, with half having ‘dot-like’ cytoplasmic desmin<br />
and WT1 positivity; some showed focal positivity for NB84, CD99 and Bcl-2.<br />
‘Undifferentiated’ sarcomas showed the widest range <strong>of</strong> staining with no<br />
marker staining all cases. Neuroblastomas exhibited uniform strong staining for<br />
CD56 and NB84 and marked cytoplasmic Bcl-2 positivity. Blastematous Wilms<br />
tumors showed uniform strong membranous staining for CD56, uniform<br />
cytoplasmic staining for Bcl-2 and nuclear expression <strong>of</strong> WT1.<br />
Conclusion. Embryonal pediatric malignancies can demonstrate apparently<br />
non-specific expression patterns for several antigens, which may reflect<br />
developmental immaturity rather than specific differentiation pathways.<br />
172<br />
Surgical Pathology <strong>of</strong> the Paediatric Larynx: A 15 Year<br />
Survey in a Specialist Children’s’ Hospital<br />
B Vadgama , A D Ramsay , M Malone , N J Sebire<br />
Great Ormond Street Hospital for Children, London, United Kingdom<br />
Examination <strong>of</strong> laryngeal specimens constitutes a significant proportion <strong>of</strong> the<br />
general adult surgical workload, and many cases involve epithelial neoplasia. In<br />
children the clinical problems and pathological lesions involving the larynx are<br />
quite different. We have examined the spectrum <strong>of</strong> laryngeal pathology seen in<br />
a tertiary referral specialist paediatric pathology unit.<br />
A total <strong>of</strong> 256 laryngeal and epiglottic specimens from 212 patients<br />
were retrieved from the pathology archives for the 15 year period 1998 to 2003.<br />
31 patients (14.6%) had more than one biopsy and the patients’ age at the time<br />
<strong>of</strong> biopsy ranged from one day to 17 years old (median 2.8 years). <strong>The</strong> male to<br />
female ratio was 1.3:1. Laryngeal specimens constituted less than 1.5% <strong>of</strong> the<br />
departmental workload.<br />
<strong>The</strong> most frequent clinical indications were laryngomalacia (18),<br />
congenital lesions (15), cysts (14) and hoarse voice (8). <strong>The</strong> majority <strong>of</strong><br />
specimens (254 cases) were biopsies and the vocal cord was the most common<br />
site. <strong>The</strong>re was a total laryngectomy for fibrotic stenosis secondary to caustic<br />
soda ingestion, and a partial laryngectomy for paraganglioma. <strong>The</strong> commonest<br />
diagnoses were papilloma (48); intubation granuloma (39); cyst (20); chronic<br />
inflammation (22); acute inflammation (10 - 3 with necrosis and 1 with fungal<br />
hyphae); and 8 each <strong>of</strong> vocal cord polyps and haemangiomas. Less common<br />
conditions included rhabdomyosarcoma (3), granular cell tumour (2), severe<br />
dysplasia (1), and lipid proteinosis (1). No abnormality was seen in 20 cases.<br />
We concluded that laryngeal pathology is relatively infrequent in<br />
the paediatric age group, but that a variety <strong>of</strong> conditions can occur at this site<br />
and that these can arise from both the epithelial and the stromal components <strong>of</strong><br />
the larynx.<br />
69