Rhabdomyosarcoma represents a large group of soft tissue sarcomas disp
laying heterogeneous histopathological features. In addition to their
histopathological classification, the variable expression of a number
of histopathological features may contribute to the heterogeneity and
may be related to prognosis. Tissue sections of 113 well-documented, p
rotocol-treated patients with long term follow-up (mean 6 years) were
analysed by a panel of four paediatric pathologists. The following fea
tures were assessed: presence of rhabdomyoblasts, degree of maturation
of rhabdomyoblasts, heterogeneous maturation patterns, mitotic activi
ty, tumour necrosis, myxoid component, and septa. A scoring system was
allocated to each index. High degree of maturation (amount of cytopla
sm greater than surface area of the nucleus), absence of tumour necros
is (< 10% of tumour surface), and absence of septa (< 10% of tumour su
rface) significantly correlated with a favourable clinical course. Rep
roducibility in the assessment of these three features was good: mean
ic varying from 0.53 to 0.64. A rhabdomyosarcoma score function for su
rvival was defined by: (- 0.27 x degree of maturation score) + (0.007
x percentage septated area) + (0.031 x percentage tumour necrosis). Ba
sed on the score a two-grade system was elaborated, i.e. grade I (scor
e < -0.20) v. grade II (score greater than or equal to -0.20). Rhabdom
yosarcoma grade appeared to be the best factor in predicting patients
survival: 69% long-term survival in patients with grade I v. 33% in pa
tients with grade II (P = 0.0001). Moreover, this grading system was s
hown to have discriminative power within the group of patients with em
bryonal rhabdomyosarcoma: patients with a grade I tumour fared signifi
cantly better than those with a grade II tumour (63% v. 41% long-term
survival, P = 0.03). This study, therefore, demonstrated the prognosti
c value of this grading system and its useful application within the g
roup of patients with embryonal rhabdomyosarcomas.