Torticollis due to contracture of the sternomastoid muscle is a well recognised entity,
but there seems little agreement as to etiology, treatment or even terminology. The deformity
is largely cosmetic and the results of treatment gratifying. The greatest interest lies in the
causation of this condition, with reference in particular tosternomastoid tumour. The
pathogenesis remains uncertain but appears unique, there being no known clinical or
histological parallel in any other disease process. The objects of this paper are to clarify the
relationship of sternomastoid tumour to torticollis, to present some observations on the two
conditions and to discuss the results of treatment.
ETIOLOGY AND PATHOGENESIS
Despite severalcomprehensive reviews, notably by Chandler and Altenberg (1944) and
Lidge, Bechtol and Lambert (1957), there is little agreement about the etiology of sternomastoid
tumour or muscular torticollis. Chandler and Altenberg regarded the progression from one
to the other as so certain that they recommended excision of all tumours at the earliest age.
By contrast, Hulbert (1950) and Coventry and Harris (1959)contended that the majority of
tumours resolve completely without active treatment. Gray (1935) and Bianco (1958) claimed
that most children presenting with torticollis give no history of a sternomastoid tumour, but
Coventry and Harris (1959) stated that in half their cases the presence of a tumour was missed
by parents and only detected on routine paediatric examination.
It is now wellestablished that a sternomastoid tumour is frequently preceded by a breech,
forceps or primiparous birth (Witzel 1883, Fitzsimmons 1933, Chandler and Altenberg 1944,
Hulbert 1950, Coventry and Harris 1959). It is tempting to assume that the common factor is
trauma. Stromeyer (1838) postulated that rupture of the muscle during parturition produced a
tumorous haematoma, and that fibrous replacementcaused subsequent torticollis. Of fifteen
neonatal or stillbirth necropsies with sternomastoid haematomata, Spencer (1893) found that
twelve were by breech or forceps delivery. Sanerkin and Edwards (1966) demonstrated extensive
post-mortem muscle fibre damage, haemorrhage and ischaemic change after a breech delivery.
Fitzsimmons (1933) reviewed fifty-four cases of Erb’s palsy and found anassociated
sternomastoid tumour in eleven. This he construed as further evidence that the tumour was
due to birth injury. Experimentally, Middleton (1930) produced histologically similar tumours
by ligation of veins draining the sartorius muscle in dogs, and inferred that venous occlusion
during parturition was the cause in babies. However, the striking histological feature of
sternomastoidtumour, described as early as 1875 by Taylor, is profuse fibrous replacement of
muscle (Figs. I and 2) without a trace of haemosiderin, even in specimens excised as early as
three weeks after birth (Chandler 1948). Reye (1951) examined four necropsy specimens at
ages between four and twenty weeks, and concluded that the appearances suggested almost
an overgrowth of tendon at the expense of muscle.Kiesewetter, Nelson, Palladino and Koop
(1955), in a study of thirty-two surgical specimens, found very small amounts of haemosiderin
in eight only. These lesions were not consistent with massive haematomata, and the tumour
much more closely resembled fibrosarcoma, fibromatosis of palmar or plantar fascia, or
desmoid tumours. Finally, birth injury cannot have been a factor in the now numerouspublished instances of sternomastoid tumour following Caesarian section (Schloessmann 1911,
Stern 1924, Rossi 1928, Chandler and Alternberg 1944, Kiesewetter et a!. 1955).
432 THE JOURNAL OF BONE AND JOINT SURGERY
FIG. 1 FIG. 2
Figure 1-Histological section from the edge of a sternomastoid tumour in a 4-week-old boy. Note
the swollen fragmented muscle fibres in the lower right hand corner,...