1. Soft Tissue Tumours in Children
Gordan M. Vujanić
School Of Medicine
Cardiff University
Cardiff
United Kingdom

2. Pathology of Tumours
The Era of Morphology - 1850-1980s
The Era of Immunohistochemistry – 1980 – 2000
The Era of Molecular Diagnosis - 2000 -

3. Tumours in Children
 Challenging to diagnose
- relatively rare
- show challenging morphological features
 Different form tumours in adults (carcinomas vs. embryonal)
 Rare even in tertiary paediatric referral centres
 Multicentre trials resulted in huge improvement (central pathology
review, biology studies, standardized multidisciplinary therapeutic
treatments)

4. Soft Tissue Tumours in Children
- WHO Classification -
1. Fibrous tissue tumours
2. Fibrohistiocytic tumours
3. Lipomatous tumours
4. Smooth muscle tumours
5. Skeletal muscle tumours
6. Endothelial tumours of blood vessels
7. Perivascular tumours
8. Synovial tumours
9. Mesothelial tumours
10. Neural tumours
11. Paraganglion tumours
12. Cartilage and bone tumours
13. Pluripotential mesenchymal tumours
14. Miscellaneous tumours
15. Unclassified tumours

5. Paediatric Soft Tissue Tumours
 Historically, classified according to their degree and type of cellular
differentiation
 Skeletal myogenesis = rhabdomyosarcoma
 Spindle cell morphology = fibrosarcoma/myofibrosarcoma
 Small round cell tumours = undifferentiated sarcoma
 Morphological diagnosis of limited value
- Tumours with similar histology – markedly different clinical
behaviour and outcome

6. Rationale for using molecular techniques
in paediatric tumour pathology
 Diagnostic difficulties (especially with needle biopsies)
 Immunophenotype often non-diagnostic
 CD56 - PNET, RMS, NB, WT
 CD99 - PNET, DSRCT, WT
 Bcl-2 - PNET, DSRCT, WT, NB, SS
 WT1 - PNET, RMS, DSRCT, NB, WT
 Important role in diagnostic and prognostic stratification of tumours
 Prognosis / treatment / trial protocols

7. Which ‘molecular’ methods are diagnostically
applicable to histopathology samples?
 Numerous techniques can be utilised in the molecular genetic
analysis of tumours
 Classical cytogenetics
 Fluorescence in situ hybridisation (FISH)
 PCR and RT-PCR

8. Paediatric Soft Tissue Tumours
 When tumour karyotyping introduced:
 STT with recurrent chromosomal translocations
- more uniform cell population
 STT with complex karyotypes lacking recurrent translocations
- more likely to be pleomorphic
- RT-PCR and FISH – sorted out many problems ass with classic
cytogenetics

9. Why molecular diagnostic techniques?
 Many paediatric tumours have specific translocations
with associated functional fusion gene products
Tumour Translocation Fusion gene product(s) Prevalence
A-RMS (2;13)(q35;q14) PAX3-FOXO1 ~70%
(2;13)(p36;q14) PAX7-FOXO1 ~10%
ES/PNET (11;22)(q24;q12) EWS-FLI1 ~85%
(21;22)(q22;q12) EWS-ERG ~10%
t(7;22)(q33;q12) EWS-FEV ~1%
DSRCT (11;22)(p13;q12) EWS-WT1 ~93%
Synovial sarcoma (X;18)(p11.2;q12.2) SYS18SSX1; SYS18SSX2; 63%; 37%
SS18-SSX4 rare
IFS (12;15)(p13;q25) ETV6-NTRK3 ~100%
DFSP (17;22)(q22;q13) COL1A1-PDGFB ~92%
AlvSPS (X;17)(p11;q25) ASPL-TFE3 ~100%

10. FISH vs. PCR in routine practice
 PCR
 Precise identification of transcripts BUT
 Limited to known targets
 Reduced sensitivity in FFPE specimens due to lower RNA quality
 Usually the housekeeping gene expressed at higher levels than target so
possible false
 FISH
 Less specific since FISH detects for example any EWS translocation
 Translocation partner NOT identified but also not necessarily required to
interpret test !

11. Role of Molecular Diagnostics in Paediatric Tumours
 As a diagnostic tool (Ewing’s sarcoma/PNET, or Synovial Sarcoma)
 Tumour Classification (Alveolar RMS vs. Embryonal RMS)
 As a prognostic marker (Neuroblastoma, RMS, etc.)

12. Soft Tissue Tumours in Children
Tumour category Children (%) Adults (%)
 Vascular 29 9
 Neurogenic 15 9
 Myogenic 14 5
 Fibroblastic-myofibroblastic 12 7
 Fibrohistiocytic 12 17
 Lipocytic 6 16
 Other 12 38

13. Rhabdomyosarcoma (RMS)
 the most common soft tissue sarcoma in childhood
 peak incidence in the first decade
 familial form is associated with Li-Faumeni syndrome with increased risk for development
of various neoplasms
 most common in soft tissues, but can occur anywhere (cutaneous RMS in association with
epidermal naevus & von Recklinghausen’s disease)

14. International Classification of RMS
I Superior prognosis Botryoid RMS
Spindle cell RMS
II Intermediate prognosis Embryonal RMS (classical)
III Poor prognosis Alveolar RMS
IV Unknown prognosis RMS with rhabdoid features

16. Diagnosis of RMS
 Morphology
 Demonstration of myogenic differentiation
 Molecular characterisation

17. Embryonal RMS
 Common in head & neck, genitourinary tract, paraspinal/parameningeal sites
 Incidence declines after age 3 yrs
 Prognosis is dependant on a number of factors:
 Site
 Age
 Staging
 Histology
 All are important and form part of an IRS grouping for patient management

18. Histopathology
 Embryonal RMS has a variable pattern
 Varying degree of cellularity with alternating dense hypercellular areas and more
loose myxoid areas
 A mixture of round cells, spindle cells and more differentiated cells showing
rhabdomyoblastic differentiation
 Anaplasia can be present in embyonal and alveolar RMS (?worst prognosis)

19. Embryonal RMS

20. Embryonal RMS

21. Alveolar rhabdomyosarcoma
 A-RMS accounts for at least 30-40% of all RMS
 Alveolar RMS displays a bimodal age incidence
 A peak at 3 years and 15 years of age
 Higher incidence in upper and lower limbs

22. A-RMS - Pathology
 Tumour composed of ill defined aggregate/nests of poor differentiated round to oval
cells
 Some cells may show rhabdomyoblastic differentiation
 The nests usually show central loss of cellular cohesion forming irregular alveolar
spaces

23. A-RMS - Pathology (continued)
 Fibrovascular septa separate the nests of cells, with single cells adherent to these
septa
 Mitotic figures are common
 Wreath like giant cells are a characteristic and diagnostically useful feature of alveolar
RMS
 The alveolar pattern is also seen in metastatic sites
 Solid variant of alveolar RMS is an important subtype in which fibrous septa may
be absent or inconspicuous

24. Alveolar RMS

25. Alveolar RMS

26. Wreath like giant cells in A-RMS

27. Solid variant A-RMS

28. Reticulin

29. A-RMS - Differential diagnosis
 Any small round cell tumour but in particular pseudo-alveolar pattern in
 PNET
 alveolar soft part sarcoma
 desmoplastic small round cell tumour

30. Pseudoalveolar pattern in PNET

31. Immunohistochemistry in RMS
Recommended panel:
Desmin (in 99%)
Myogenin (in 100%) – most specific
MyoD1 (in 100%)
Muscle Specific Actin (in 94%)
CD99 (weak granular) (in 14%)
CAM5.2
EMA
Synaptophysin
Other useful markers
Myoglobin (in 78%) in more differentiated forms
Sarcomeric actin
Myosin

32. Desmin staining

33. Myo D1 nuclear staining

34. Myogenin CD99

35. Rhabdomyosarcoma -
immunohistochemistry vs prognosis
 Myogenin - >80% ARMS vs >25% ERMS (>50% of tumour cells)
 Diffuse myogenin + a marker of poor prognosis (independent of fusion
status and histology)
 PAX-5 – 67% ARMS vs 0% ERMS

36. Rhabdomyosarcoma -
molecular biology
 ERMS associated with multiple numeric and structural chromosomal changes (extra
copies of chromosomes 2, 8 and 13; LOH at 11p15 – BW region)
 ARMS associated with two specific gene fusions
 PAX3-FOXO1 (correlated with t(2;13)(q35;q14) – 3x more common
 PAX7-FOXO1 (correlates with t(1;13)(p36;q14) – better prognosis
 20-25% ARMS have neither of these translocations – their genetic profile
identical to ERMS - ?prognosis

37. Synovial Sarcoma
 5-10% of STS
 42% of paediatric non-RMS STS
 85-95% arise in the extremities, then head and neck
 Biphasic pattern – spindle cells and epithelial gland components
 In children – monophasic SS more common

39. Focal epithelial differentiation

40. Cytokeratin staining

41. SS - Immunoprofile
 Cytokeratin + (some are -)
 EMA +
 Vimentin +
 S100 + in 30% of cases
 CD99 + in 60% of cases
 BCL2 + in 75% - 100% of cases
 CD34 -
 Calponin +
 Desmin -
 TLE1 - nearly 100%

42. Synovial Sarcoma
 Characterised by a translocation: t(x;18)(p11.2;q11.2)
(SS18-SSX1); (SS18-SSX2); (SS18-SSX4)
 The identification of a specific translocation has resulted in the ability to
diagnose these tumours in unusual sites (viscera)
 It is also possible to identify poorly or undifferentiated SS
 Is an important tool in differentiating SS from other spindle cell tumours
(congenital fibrosarcoma)

43. Congenital Infantile Fibrosarcoma
 Diagnosed in the 1st year of life (~50% congenital)
 Usually occur in distal extremities & head and neck
 Histologically simulates adult-type fibrosarcoma
 5-year survival of >90%
 Recurrence rate 30%
 Metastases rare
 Chromosomal translocation t(12;15)(p13;q26) (ETV6/NTRK3 genes)

44. Congenital Fibrosarcoma

45. Vimentin staining

46. EWS/PNET
Ewing Family of Tumours
 EWS/PNET - a group of tumours with overlapping clinical & pathological
features
 peak incidence in childhood & young adolescence
EWS – classically a primary bone neoplasm (rarely extra-osseous)
PNET – usually arises in soft tissues (exceptionally in bone)

47. PNET - Pathology
 i) lobular, ii) diffuse & iii) incohesive growth pattern
 foci of haemorrhage but NO calcification
 tumour lobules divided by fibro-connective septae
 Homer-Wright & perivascular pseudo-rosettes rare
 small cells, round-oval nuclei, coarse chromatin, small nucleoli, narrow rim of
eosinophilic cytoplasm
 intercellular fibrillary material or ganglion cells exceptionally rare

50. IHC of EWS/PNET – essential
POSITIVE for:
 CD99 ~95%, perimembranous
 VIM ~95%, focal or diffuse, dot-like
or perinuclear
 NSE, S100, synaptophysin, NFP ~40%
 Chromogranin ~20%
 GFAP <10%
 EMA, Cytokeratin, Desmin <10%

51. CD99

52. Ewing’s Family of Tumours –
molecular biology
 t(11;22) (involving EWS gene) – 85-90% of ESF
- t(11;22)(q24;q12) (EWSR1-FLI1)
- Type 1 (exon 7 – exon 6) - ?better prognosis
- Type 2 (exon 7 – exon 5)
- t(21;22)(q22;q12) (EWSR1-ERG) – 5-10% of ESF
- 4-9% of ESF – other fusions

53. Desmoplastic Small Round Cell Tumour
- DSRCT
 rare, highly aggressive tumour of uncertain origin
 M>F
 presents with abdominal distension, pain & mass
 at laparotomy – variable sized mass with numerous smaller nodular peritoneal
implants throughout the peritoneal cavity

55. DSRCT -
Immunohistochemistry
Not specific but distinct co-expression of
i) epithelial (AE1/AE3, CAM5.2)
ii) neural (NSE, PGP9.5; CD99 in 20-35%)
iii) muscle markers (desmin; negative for MyoD1, myoglobin)

56. DSRCT – Molecular biology
 This tumour is defined by a reciprocal translocation
t(11;22)(p13;q12) (EWS-WT1 genes)
 This translocation creates a fusion gene that transcribes a chimeric
protein containing portions of WT1 protein & EWS protein

57. DSRCT - Prognosis
 a very aggressive neoplasm
 chemotherapy may induce temporary remission
 the ultimate outcome remains dismal

58. Molecular findings are the Gold Standard for
diagnosis ?
 ‘Non-specificity’ of immunomorphology
 CD56 - PNET, RMS, NB, WT
 CD99 - PNET, DSRCT, WT
 Bcl-2 - PNET, DSRCT, WT, NB, SS
 WT1 - PNET, RMS, DSRCT, NB, WT
 vs. specific diagnosis with molecular techniques
 ETV6/NTRK3 in Infantile fibrosarcoma
 SS18-SSX1 in Synovial sarcoma
 EWS-FLI1 / other translocations in atypical PNET

59. Histomorphology is the Gold Standard for
diagnosis?
 ‘Non-specificity’ of molecular markers
 12;15 ETV6-NTRK3 in CMN/IFS and secretory breast
carcinoma
‘A FISH study of ETV-NTRK3 fusion gene in secretry breats carcinoma’
Marketsov et al. Genes Chrom Cancer 2004; 40: 152-7
 X;18 in SS and MPNST
‘Lack of SYT-SSX fusion transcripts in malignant peripheral nerve sheath tumors on RT-
PCR analysis of 34 archival cases’ Tamborini et al. Lab Invest 2002; 82: 609-18
 Xp1 RCC and AlvSPS
‘Primary renal neoplasms with the ASPL-TFE3 gene fusion of alveolar soft part sarcoma: a
distinctive entity previously included among renal cell carcinoma of children and
adolescents’
Argani et al. Am J Surg Pathol 2001; 159: 179-92

60. SUMMARY
 Molecular studies invaluable for diagnostic and clinical management of a range of
tumours
 In the future, clinical decisions will increasingly be based on a combination of
histological criteria and the molecular identification of genetic abnormalities
 Future advances likely to lead to even greater diagnostic and prognostic information,
and identification of new therapeutic targets in treatment
 Gold standard is interpretation of molecular findings in the context of the
histopathological features