Streptococcal necrotizing myositis is a rare but severe infection
of skeletal muscle that is caused almost exclusively by Streptococcus
pyogenes (Group-A streptococcus). The literature on this
condition is sparse. In 1930, Abrami and Worms1 reported
on two patients who died of this infection. Since then, fewer than
thirty cases have been reported, to our knowledge2-9.
Streptococcal necrotizing myositis has a much more sinister prognosis
than do other bacterial infections of skeletal muscle, such as pyomyositis.
The condition is difficult to diagnose; hence, treatment is often
delayed, contributing to its associated high mortality rate.
We report a case of a five-year-old girl who survived streptococcal
necrotizing myositis. A high index of suspicion combined with emergent
magnetic resonance imaging facilitated early diagnosis and radical
débridement of the necrotic muscles. The clinical and magnetic
resonance imaging features of streptococcal necrotizing myositis
are compared with those of pyomyositis.
A five-year-old nondiabetic girl presented with a twenty-four-hour
history of severe pain in the left hip and thigh following a single
day of fever that had resolved spontaneously one week prior to admission.
There was no recent history of injury or injection in the region
of the left buttock, hip, or thigh; chicken pox; or steroid intake.
The patient was attached to a cardiac monitor, and blood pressure
was checked every fifteen minutes. Although she was febrile (39.5°C)
and dehydrated, her vital signs were initially stable (pulse, 120
bpm; blood pressure, 100/50 mm Hg). The pain in the left
lower limb prevented walking, and muscle spasm markedly limited
the range of motion of the hip. The hip could be flexed passively
to only 20°; other directions of movement were virtually impossible.
Although no abnormalities of the overlying skin were evident, the
left hip and thigh were diffusely swollen, with exquisite tenderness
without crepitus on deep palpation over the buttocks.
A conventional anteroposterior pelvic radiograph revealed marked
soft-tissue swelling around the left hip but no soft-tissue gas.
A diagnosis of infective myositis that was more likely to be streptococcal
necrotizing myositis because of the toxic clinical presentation
was made. The patient was prepared for surgery and an emergent magnetic
resonance imaging scan. Blood was drawn for culture prior to administration
of amoxicillin-clavulanate and clindamycin. Additional investigation revealed
a white blood-cell count of 22.3 109/L
(leukocytosis), a hemoglobin level of 120 g/L, and a platelet
count of 321 109/L; both the
erythrocyte sedimentation rate and the C-reactive protein level
were elevated (to 80 mm/hr and 20 mg/L, respectively).
The prothrombin time (16.6 sec), activated partial thromboplastin
time (49.5 sec), and international normalized ratio (1.6) were all
prolonged.
A magnetic resonance imaging scan, performed emergently within
one hour after admission, demonstrated marked edema of the left
gluteus medius and minimus muscles without associated articular
or osseous abnormality (Figs. 1 and 2). Vital signs were monitored throughout
the procedure, and they remained stable. General anesthesia was
not required.
Following magnetic resonance imaging, the patient’s
condition deteriorated and she became hypotensive (blood pressure, 70/30
mm Hg), requiring vasopressor support. Surgery was performed under
general anesthesia within two hours after admission, through a lateral
approach to the left hip. The gluteus maximus appeared normal, but
the gluteus medius and gluteus minimus were grossly swollen and
a dull red color. Normal contractility and capillary bleeding of
these two muscles were absent. No abscess was found. A radical excision
of the gluteus medius and gluteus minimus was performed. Neither
arthrotomy nor aspiration of the left hip joint was performed as
the hip had appeared normal on magnetic resonance imaging. Following
débridement, the patient’s blood pressure improved
slightly to 85/45 mm Hg, and she was transferred to the
intensive-care unit, where she required ventilator and vasopressor
support.
The blood pressure was maintained at 90/45 mm Hg during the
first two postoperative days. Two days later, exploration was performed
to determine whether repeat débridement was needed, but
it revealed no additional tissue necrosis. The wound was closed
in layers. Vasopressor support was gradually reduced and was finally
discontinued on the fourth postoperative day. Also on that day,
the patient’s temperature returned to normal and she was
extubated. She was fully conscious, and she was able to tolerate
an oral diet by the fifth postoperative day. The wound healed well,
and the sutures were removed on the fourteenth day.
A pure growth of Streptococcus pyogenes that
was sensitive to penicillin G and clindamycin grew on culture of
muscle specimens and blood. Specimens obtained with swabs from the
throat and nose of the patient and the other family members were
negative for Streptococcus pyogenes. Histological
examination of the excised muscles showed extensive tissue necrosis
and heavy suppurative inflammation, mainly affecting the fibrous
septa between the muscle bundles and in the fascial tissue.
In contrast to pyomyositis, an infection of skeletal muscle that is
caused predominantly by Staphylococcus aureus and
is associated with a mortality rate of <1%10,11, streptococcal necrotizing myositis
is a rare but much more severe infection with a mortality rate between
50% and 100%2-9.
Although in the literature this severe infection has usually been
referred to as streptococcal myositis, we prefer the term streptococcal
necrotizing myositis because extensive muscle necrosis is common,
even in the early stages of the disease12.
Infective myositis, including streptococcal necrotizing myositis
and pyomyositis, poses a diagnostic challenge to the clinician6. The symptoms and signs of myositis
can be nonspecific and misleading, not clearly revealing the involvement
of deep skeletal muscle. The skin overlying the infected muscles
usually appears normal until a very late stage. This "silent" external
appearance is in contrast to the severe pain and exquisite tenderness
of the affected area. Pain that is out of proportion to the physical
appearance has been typically described with this condition6,10,13. The clinical presentation
may be confused with other conditions, such as septic arthritis
and deep-vein thrombosis. In this five-year-old girl, the severe
pain and tenderness, the global decrease in the range of motion
of the left hip, and the diffuse soft-tissue swelling made a diagnosis
of septic arthritis unlikely. These physical findings, together
with the toxic appearance of the patient and the findings on the
anteroposterior radiographs, resulted in the diagnosis of streptococcal necrotizing
myositis.
It is important to differentiate streptococcal necrotizing myositis
from pyomyositis because urgent surgical exploration can be lifesaving
for the patient with streptococcal infection, whereas more conservative
methods can be used for the patient with pyomyositis. Clinically,
the early and rapid development of severe pain and tenderness is
more suggestive of streptococcal necrotizing myositis. The rapid
downhill course is in contrast to that seen in pyomyositis, which
can take up to three weeks to progress from the initial "invasive
stage" to the "late stage."10 The early development of toxic systemic
involvement (streptococcal toxic-shock syndrome14)
is typical of streptococcal necrotizing myositis. It is, however,
important to note that children are able to compensate well initially
but that their condition may suddenly deteriorate.
In the current case, the involvement of muscles was seen on the
plain radiographs. To delineate the exact extent of muscle involvement,
we performed an emergent magnetic resonance imaging scan. While
other imaging techniques such as computed tomography and ultrasonography
are more readily available and can provide similar information more
quickly15, their smaller field
of view does not give the surgeon as extensive a view of the infected
area. The interpretation of ultrasound images is also more operator-dependent.
Moreover, the use of a scanning probe on an exquisitely tender area is
usually unacceptable to the patient, especially a child. The advantages
of magnetic resonance imaging compared with computerized tomography
include the ability to perform multiplanar imaging and to obtain
better resolution and contrast in delineating soft-tissue lesions16,17.
The involvement of two adjacent muscles, seen on magnetic resonance
imaging, is more suggestive of an infectious etiology than of other
causes of myositis, such as hemorrhage or infarction. Magnetic resonance
imaging also helps to differentiate streptococcal necrotizing myositis
from pyomyositis. The magnetic resonance imaging criteria for the
diagnosis of pyomyositis are well documented11.
In our patient, the magnetic resonance imaging features of the involved
glutei were different from those described for pyomyositis. Unlike
in pyomyositis, liquefaction of the infected muscles to form an
abscess is not a feature of streptococcal necrotizing myositis.
The magnetic resonance imaging findings of pyomyositis, such as
rim enhancement of the infected area on contrast-medium-enhanced
T1-weighted images and high-signal-intensity fluid collection on
T2-weighted images, are therefore absent on magnetic resonance images
of streptococcal necrotizing myositis. Adjacent reactive joint effusion,
another common finding on magnetic resonance images of patients
with pyomyositis, was also absent in our patient, probably because
of the much shorter duration of the disease process. The emergent
magnetic resonance imaging also helped to exclude other diagnoses,
especially septic arthritis. In the absence of magnetic resonance
imaging, arthrocentesis might be required before the diagnosis of
septic arthritis can be confidently excluded. During performance
of this procedure, however, the virulent Streptococcus pyogenes might
be introduced into the sterile hip joint.
A prompt and radical débridement is the key to the survival
of a patient with streptococcal necrotizing myositis7. We recommend surgical exploration
after the diagnosis has been made. The magnetic resonance imaging
findings assisted us in achieving a complete, radical débridement;
because they clearly showed that the left hip joint was not involved,
exploration of the joint was deemed unnecessary. The extent of the débridement
must be guided by both magnetic resonance imaging and intraoperative
findings. The gluteus medius and gluteus minimus were found to be
nonviable intraoperatively, thus confirming the extent of involvement
seen on the magnetic resonance images. The first débridement
was adequate, and no further muscle necrosis was noted during the
second exploration. Therefore, magnetic resonance imaging provided a
reliable "roadmap" and gave the surgeons a good
mental picture for the débridement.
Despite the advantages of using magnetic resonance imaging in
patients with this severe form of infection, we must stress that
emergent magnetic resonance imaging may not be as readily and widely
available as other imaging modalities. Moreover, magnetic resonance
imaging usually takes more time than does ultrasound or computed
tomography, and time is an important consideration with regard to
this condition, especially if the patient is critically ill. If
magnetic resonance imaging is performed, close monitoring of the
patient’s vital signs is essential during the procedure.
In conclusion, we recommend that a high index of suspicion be
maintained and a diagnosis of streptococcal necrotizing myositis
be entertained for a patient with unexplained severe muscle pain
and soft-tissue swelling accompanied by shock. The value of emergent
magnetic resonance imaging, if it is readily available, is that
it can confirm the clinical suspicion of myositis and exclude other
diagnoses. It provides a reliable "roadmap" for
débridement because it can clearly delineate the extent
of the infection. Although no investigation can replace a high index
of suspicion and repeated meticulous physical examination, emergent
magnetic resonance imaging facilitated an early diagnosis and a
thorough radical surgical débridement in our patient, both
of which contributed to saving her life.