Acute Myopathy of Intensive Care:

David Lacomis, MD,*t Michael J. Giuliani, MD,* Anne Van Cott, MD,* and David J. Kramer, MDS

An acute myopathy of intensive care occurs in critically ill patients treated with intravenous corticosteroids and neuro-
muscular junction-blocking agents. The full clinicopathological spectrum is uncertain. We evaluated the clinical, elec-
trodiagnostic, and histopathological features of 14 patients who developed acute myopathy of intensive care after organ
transplantation or during treatment of severe pulmonary disorders and sepsis. Patients received high-dose intravenous
corticosteroids, usually in conjunction with relatively low to moderate doses of neuromuscular junction-blocking agents.
After discontinuation of the latter drugs, most had difbse, flaccid weakness with failure to wean from mechanical
ventilation. Electrodiagnostic findings were consistent with a necrotizing myopathy. Muscle histopathology revealed
myopathy with loss of thick filaments in 79%, mild myopathic changes in 14%, and atrophy of type 1 and type 2
fibers in 7%. Loss of thick filaments was identified in muscle biopsy specimens obtained 30 f 11 days (mean f standard
deviation) after intravenous corticosteroid treatment but not in those obtained earlier (12 f 2 days). Critically ill
patients, including those receiving organ transplants, may develop acute myopathy of intensive care after exposure to
intravenous corticosteroids and neuromuscular junction-blocking agents, although the exposure to the latter drugs may
be minimal. Selective loss of thick filaments is common in acute myopathy of intensive care, especially if the muscle
biopsy specimen is obtained 2 weeks or more after intravenous corticosteroid exposure.
Lacomis D, Giuliani MJ, Van Cott A, Kramer DJ. Acute myopathy of intensive care: clinical,
electromyographic, and parhological aspects. Ann Neurol 1996;40:645-654

During the last two decades, neuromuscular disorders
were recognized as common causes of weakness oc-
curring in critically ill patients. Bolton, Zochodne, and
their colleagues [ 1-41 provided extensive evidence that
an axonal sensorimotor polyneuropathy, termed critical
illness polyneuropathy (PN), frequently affects patients
who receive a week or more of treatment for sepsis or
multiorgan failure. Prolonged neuromuscular junction
(NMJ) blockade was identified in occasional intensive
care unit (ICU) patients who had persistent weak-
ness after prolonged use of neuromuscular junction-
blocking agents (NMBAs) [5-71. Although a myo-
pathic process was identified in some of the earlier
reports of weak ICU patients [5, 8, 91, an acute myo-
pathic syndrome of critical illness was not identified as
a common entity for another 5 to 10 years [lo-361.
This acute myopathy of intensive care (AMIC) is
associated with the use of intravenous (IV) corticoste-
roids (CSs), nondepolarizing NMBAs, or both. The
myopathy was first systematically documented in pa-
tients treated intensively for status asthmaticus [8-23,
25-28], but patients treated for other severe pulmo-
nary disorders and critical illnesses including sepsis and
burns were also affected [24, 29-32, 34, 351. Afflicted
patients develop acute, diffuse, flaccid weakness often
associated with failure to wean from mechanical venti-
lation. Elevations in serum crearine kinase (CK) con-
centration are noted in many. Following discontinua-
tion of the CSs and with resolution of the underlying
critical illnesses, the myopathy improves over weeks to
months.
Despite thorough descriptions of the clinical features
of AMIC, the exact pathophysiological mechanisms of
this presumed toxic process remain uncertain. Elec-
trodiagnostic studies are consistent with a necrotizing
myopathy, although a component of terminal motor
axonopathy is difficult to exclude [27,30, 3 11. Concur-
rent evidence of prolonged NMJ blockade is occasion-
ally identified, especially early in the course of the
illness [25, 321. Perhaps the most important data re-
garding pathogenesis come from the pathology studies.
In some patients with AMIC, histopathology reveals a
myopathy with a selective patchy loss of thick (myosin)
filaments [18, 20, 25, 29, 32, 361. This histopathology

From the Departments of *Neurology, tPathology (Neuropathol-
ogy), and $Anesthesia and Critical Care (Liver Transplantation
Critical Care Unit), University of Pittsburgh School of Medicine,
Pittsburgh, PA.
Received Mar 18, 1996, and in revised form May 2. Accepted for
publication May 2, 1996.
Address correspondence tO Dr Lacomis, Departmenr of ~ ~ ~
University of Pittsburgh, 322 Scaife Hall, PittAburgh, PA 15261.

Copyright 0 1996 by the American Neurological Association 645

is reproduced in an animal model utilizing denervation
a n d CS administration. I n this model, only the combi-
nation of nerve trauma and CSs causes reversible loss
of thick filaments [37, 381. In biopsy specimens from
other patients with AMIC, reported histopathological
findings are less distinctive a n d range from type 2 fiber
atrophy to necrosis [9, 11-13, 17, 19, 21-23, 26, 29-
31, 341.
M a n y reports of AMIC emphasized the clinical a n d
electrodiagnostic abnormalities b u t were limited by a
lack of pathological correlation. T h u s , the full range of
clinicopathological features is uncertain. In this article,
we describe 14 patients with AMIC who underwent
detailed clinical, electrophysiological, a n d pathological
evaluations.
Materials and Methods
Patients
During an 1 1-month period, there were approximately 8,000
ICU admissions and transfers into our ICUs. Sixteen ICU
patients were diagnosed with AMIC. Critical care staff usu-
ally referred patients because of moderate to severe muscle
weakness or failure to wean from mechanical ventilation.
Two patients refused muscle biopsy and were excluded from
this report. The other 14 were included retrospectively. In
comparison, 5 ICU patients with acute weakness were diag-
nosed with critical illness PN during the study period. No
patient with persistent weakness from prolonged NMJ block-
ade was identified.
The AMIC patients’ medical records were reviewed.
Drugs that were administered, including doses for CSs and
NMBAs, were recorded. Laboratory results from the time of
ICU admission until the time of muscle biopsy were noted
and included complete blood cell count (CBC); values for
electrolytes, glucose, calcium, magnesium, phosphorus, albu-
min, CK, creatinine, and liver enzymes; and microbiological
assay findings. In general, chemistries and CBC were assessed
nearly daily, while albumin and liver enzymes were mea-
sured less frequently. With rare exceptions, CK was first as-
sessed after neurological evaluation and prior to electromyog-
raphy (EMG).
Manual muscle strength testing was performed on 16 to
20 muscle groups, 1 or 2 days before or coincident with
EMG, by a neurologist and strength was graded on a 10-
point modified Medical Research Council (MRC) scale. A
mean MRC score (range, 0-10) (Table 1) represents the
average score of all muscles examined [39].
Diagnostic Studies
Nerve conduction studies were performed at the bedside with
a Nicolet Viking I1 unit using routine percutaneous stimula-
tion. Limbs were warmed with heat packs if the hand tem-
perature was less than 33°C or if the foot temperature was
less than 32°C. The nerves studied are listed in Table 2.
Baseline 3-Hz repetitive nerve stimulation was performed on
either the ulnar or the median nerve. If the patient was able
to exercise for 10 seconds, postexercise stimulation (single
shock) was performed on a motor nerve that was not exam-
ined by repetitive stimulation. In most cases, if the patient
646 Annals of Neurology Vol 40 No 4 October 1996
could not exercise, repetitive stimulation at 20 to 30 Hz was
performed. Concentric needle EMG was performed on distal
and proximal upper- and lower-extremity muscles.
Muscle biopsies were performed at the bedside via percuta-
neous needle biopsy using the 14-gauge Biopry system and
the method of Cot; and colleagues [40]. Biopsy of the vastus
lateralis was performed in Patient 14 and of the deltoid in all
others. Percutaneous biopsies were performed because many
patients, their families, and ICU physicians are often reluc-
tant to allow us to obtain muscle tissue by open biopsy in
the critical illness setting. By obtaining multiple samples, we
were able to examine at least 450 myofibers (usually >
1,000).
Cryostat sections (8 pm), embedded in gum tragacanth,
were stained with henutoxylin and eosin (H&E), Gomori
trichrome, NADH-TR, ATPase at pHs 4.3, 4.6, and 9.4,
oil red 0 , and periodic acid-Schiff (PAS). Sections from
6 patients were reacted for actin (Sigma) at a dilution of
1 : 150 and pan-myosin (monoclonal antibody against fast
and slow heavy-chain isoenzymes, Amersham Life Science)
at a dilution of 1 :5 via peroxidase immunohistochemistry.
At least nine fascicles (range, 9-49) containing 50 or more
muscle fibers were available for review for each patient. Paraf-
fin sections, processed from specimens from 11 patients, were
stained with H&E. Tissue was fixed in glutaraldehyde at the
time of biopsy and processed routinely for electron micros-
copy (EM).
Muscle fiber diameters from 200 fibers were measured us-
ing an eyepiece graticule on ATPase-reacted cross sections.
If the section was slightly oblique instead of transverse, the
minimum diameter was measured. The approximate percent-
age of fibers undergoing degeneration or regeneration or ex-
hibiting irregular or absent ATPase staining was judged sub-
jectively on examination of at least 500 myofibers (Table 3).
Statistics
The Wilcoxon rank test (two tailed) was used to determine
whether there was a difference in the rank totals regarding
the times when the muscle biopsies were performed between
the two groups of patients with muscle histopathological
findings that did and did not include loss of thick filaments.
Results
Clinical and Laboratory Features
T h e sex, age, cumulative IV CS a n d NMBA doses, CK
levels, a n d major diagnoses for each patient are listed in
Table 1. Two of the patients with chronic obstructive
pulmonary disease were admitted for lung volume re-
duction surgery for end-stage emphysema. Four of the
liver transplantation patients h a d fulminant hepatic
failure at the time of transplantation; the other 3 had
chronic liver disease. All 7 developed weakness soon
after receiving a n orthotopic liver transplant.
T h e transplant patients were n o t septic w h e n my-
opathy was detected. T h e nontransplant patients fre-
quently h a d multiple medical problems leading to ICU
admission including sepsis. All patients received multi-
ple drugs including antibiotics. T h e nontransplant pa-
tients who did not have bacteremia had a focus of
Table 1. Summary of Clinical and Laboratory Features
NMBA:
mg of
IV MP, VeclPan
Days from in mg (No. of MRC
Patient Age (yr)/ IV M P until (No. of Rx Vent C K (<200 Score
No. Sex Disorders Muscle Biopsy Rx Days) Days) Dep IWlliter) (0-10) Outcome

1 62/M Severe C O P D ; slp bul- 47 3,180 26210 X 236 0 Died

lectomy; ischemic co- (27) (8)
litis; sepsis
2 581M Liver transplantation; al- 39 3,615 26"Ilo" X <20 2 Walked with assis-
coholic cirrhosis (19) (1) tance at 3 mo
3 44lM Pneumonia; ARDS; his- 24 1,760 145185 X 219 0 Ambulated with
tory of alcoholism; re- (9) (10) AFOs in 4 mo
nal insufficiency
4 66lM Severe COPD; s / p bul- 42 1,590 21"O X <20 3 Transfer to ventila-
lectomy (17) tor facility at 2
mo; no further
follow-up
5 621M slp 2 Liver trans- 18 3,240 90124 X 331 0 Ambulatory at 1
plantations: sclerosing (16) (3) mo; normal in
cholangitis 2 mo
6 651M s l p 2 Liver trans- 32 3,930 90124 X 22 0 Multiple complica-
plantations; hepatitis (10) (5) tions; ambula-
C; diabetes tory at 12 mo
7 6llF Liver transplantation; 20 1,760 7113",b X <20 4 Ambulated well in
autoimmune hepatitis (1 1) (1) 2 mo
8 791F Small-bowel obstruc- 31 432 20"lO 0 <20 8 Lost to follow-up
tion; s l p lysis of adhe- (10)
sions; COPD; pulmo-
nary edema; Staph.
aureus sepsis
9 57lM Severe COPD; pulmo- 14 2,320 14410 X 136 0 Died on day 30
nary aspergillosis; s l p (8) (4) from sepsis and
thoracotomy; gastro- hemorrhage
intestinal bleeding
10 65lF COPD; pneumonia; re- 35 1,520 010 x 125 6 TO rehab after 1
spiratory failure (9) mo; no further
follow-up
11 461F slp Lung transplanta- 31 1,635 30/30" X 92 2 Ambulatory in 3
tion; bronchoalveolar (13) (2) mo
cancer
12 49lM Liver transplantation; 10 1,620 16122 X <20 2 Ambulated with
hepatitis C; alco- (10) cane at 6 wk
holism
13 65lF Liver transplantation; 10 2,480 10139" X 103 1 Mild proximal WE
hepatitis C (10) (1) weakness and
moderate LE
weakness at 2
mo
14 461M Liver transplanration; 12 3,220 0118' 0 <20 5 Ambulated in 1
graft rejection; as- (5) wk
cending cholangitis;
history of sclerosing
ch o1an git is

'Administered in operating room only.

bDoxocurium rather than pancuronium.
IV M P = intravenous methylprednisolone; Rx = treatment; NMBA = neuromuscular junction-blocking agent; Vec = vecuronium; Pan
= pancuronium; Vent Dep = ventilator dependent; C K = creatine kinase; M R C = Medical Research Council; X = present; 0 = absent;
AFO = ankle foot orthoses; UE = upper extremity; LE = lower extremity; C O P D = chronic obstructive pulmonary disease; s/p = status
post; ARDS = adult respiratory distress syndrome.

Lacomis et al: Myopathy of Intensive Care 647

Table 2. Elertrodiagnostic Studies

Needle Examination
Nerve Conduction Studies: Amplitudes (pV for Sensory; mV for Moror)
Patient Fibs UE Fibs LE SDILAIP
No. Surd S Med S Rad S Ulnar S Per M Tib M Ulnar M Med M ProxlD ProxiD MUPs
~ ~~

1 10 11 12 5 0.8" 2b NR" NR +I+ +I+ NF

2 5 27 14 0.3' 4 2a +/+ +I+ +lo/+
3 6 18' 5" 11' 0.2" 3.6" 1" la 01 + +/+ +/+I0
4 7 13 0.5 6 +I+ +/+ +/+I+
5 NK" 9 1' 13 +I0 +I+ +/+I+
6 5h 8' 5 2" 2^ + /0 01 + +I+/+
7 3 41 1.6" 3" +/0 + /0 +/+/+
8 10 14 9 I' 3" 2= 010 010 +/+/+
9 2h 15 1 .3a 3b +/+ -1- +/+I+
jI, 4['
10 7 36 9' 15 NR" 2 3 +I+ +I+ +/+I+
11 6 26 27 ld 5b 2" +I+ +/+ +lo/+
12 3 13b 6" 8b 0.7' 4b 3" 2" +/NA +/NA NF
13 6 18 14 14 2.6 5 6' 4' 010 010 +/0/ +
14 NR (E)" 12b 8b 0.3d 8 10 010 010 +lo/ +
"educed more than 50%
"Reduced 21 -50%.
'Reduced 1-20% for age.
M = motor; S = sensory; Med = median; Rad = radial; Per = peroneal; Tib = tibial; Fibs = fibrillation potentials; UE = upper extremity;
LE = lower extremity; Prox = proximal muscles; D = distal muscles; SD = short duration; LA = low amplitude; P = polyphasic; MUPs
= motor unit potentials; N R = no response; E = edema; NF = no firing motor unit potentials; NA = not available; + = present; 0 =
absent.

infection (most often pulmonary), leukocytosis, and
usually hypotension or multiorgan failure.
All patients had hyperglycemia that required inter-
mittent insulin administration. The serum glucose level
was periodically higher than 300 mg/dl (normal, 70-
110 mg/dl) in 13 of the 14 patients. Three patients
had renal insufficiency (creatinine >2.0 mg/dl) that
did not require dialysis; 4 (Patients 3, 6, 12, and 13)
received hemodialysis for overt renal failure. The non-
transplant patients had normal liver function, except
for Patients 1 and 3 who had moderate elevations in
liver enzyme levels. Serum magnesium concentrations
were generally normal. Transient hypophosphatemia
occurred in 9 patients. Intermittent mild hyponatremia
and hypernatremia were common. Albumin levels were
low, at least transiently, in all patients. All patients
were anemic.
Patients were noted to be weak 4 to 40 days (mean,
16 days) after ICU treatment, which included intrave-
nous CSs, NMBAs, or both. Since patients received
NMBAs and sedatives, the exact time of onset of weak-
ness was usually not possible to determine. Weakness
was generalized and always affected neck flexors. Distal
and proximal muscles were affected approximately
equally in 9 patients, and proximal muscles were
weaker in 5 (Patients 2, 7,8, 10, and 11). Facial mus-
cles were weak in 8 patients (57%); only Patient 11
(7%) had extraocular muscle weakness. Eleven patients
were severely weak (MRC score <5 of 10); the others
had more moderate weakness but they were unable to
ambulate. Twelve patients (86%) were slow to wean
from mechanical ventilation. Three patients (Patients
I , 3, and 6) were areflexic, 4 had hyporeflexia (Patients
5, 8, 10, and 11), 5 had normal tendon reflexes except
for absent ankle jerks (Patients 2, 4, 7,12, and 14),
and 2 had normal reflexes including ankle jerks (Pa-
tients 9 and 13). All patients at least grimaced to pain,
but sensory testing was initially unreliable in most. No
patient reported dysesthesias during the recovery pe-
riod. Four patients (Patients 2, 4, 9, and 12) were tran-
siently encephalopathic.
Because of severe multisystem disease, 2 patients
were allowed to die. Postmortem evaluations were not
allowed. In most of the others in whom follow-up was
available, strength improved such that unassisted am-
bulation occurred within 3 months.
Electrodiagnostic Testing
Motor and sensory amplitudes are listed in Table 2.
Conduction velocities were normal except for mild
slowing in Patient 12. Excluding Patient 12, distal la-
tencies were normal except for isolated median prolon-
gation due to carpal tunnel syndrome (3 parients). Re-
petitive stimulation revealed no significant decremental
(>10%) or incremental responses.
Needle examination revealed a variable degree of

648 Annals of Neurology Vol 40 No 4 October 1996

Table 3. Histopathological Features
Mean Type 1 Fiber Mean Type 2 Fiber Irregular
Patient Diameter (pm) Diameter (pm) ATPase EM: Thick
No. (Normal, 53-61) (Normal, 42-62) Necrosis Regeneration Staining Filament Loss
1 17" 17" ++ ++ +++ X
2 25 10 + + + X
3 26 17 0 + ++ X
4 32 33 + + ++ X
5 40 32 + + ++ X
6 37 25 + + + x
7 39 30 + 0 + X
8 47 20 + + + X
9 51 38 + 0 0 0
10 24 18 ++ + + X
11 47 28 + + ++ X
12 48 33 0 0 0 0
13 46 29 0 + + X
14 52 36 + + 0 0
'Cannot differentiate fiber types due to very irregular ATPase staining and necrosis.
EM = electron microscopy; 0 = not present; + = in 1-33% of myofibers; ++ = in 34-65%; +++ = in 66-99%; ++++ = in
100%; X = present.

often diffuse fibrillation potential activity in 11 (79%)

of the 14 patients. Patient 1 also had fibrillation poten-
tials in facial muscles. Early or sometimes normal re-
cruitment of motor unit potentials (PAUPS)was noted.
Some patients could not activate any MUPs in individ-
ual muscles that were severely weak. Short-duration,
polyphasic, and often low-amplitude MUPs were noted
in 13 patients. Patient 1 had no firing MUPs in limb
muscles but had very polyphasic MUPs in facial mus-
cles. Patient 12 had a limited study of only two muscles
due to severe thrombocytopenia and an inability to ac-
tivate any muscles; a follow-up study was refused. Serial
studies were not performed.

Muscle Pathology
Moderate to severe type 2 fiber atrophy as well as a
mild to moderate reduction in mean type 1 fiber diam-
eters was evident in all biopsy specimens (see Table
3). The atrophic fibers were generally angulated and
sometimes occurred in small groups. However, fiber-
type grouping was only evident in the specimens from
Patients I and 12. Myofiber degeneration, regenera-
tion, or both were noted in 13 of the 14 patients.
Rimmed vacuoles were noted in one specimen (Patient
7). Abnormal myosin-ATPase staining (at all pHs) oc-
curred in some nonnecrotic fibers in 11 specimens
(78%) (Fig 1).
The pattern of decreased ATPase staining was
(a) patchy within atrophic and nonatrophic fibers of
both fiber types (seven specimens), (b) diffuse but only
within scattered atrophic type 1 and type 2 fibers (one
specimen), and (c) patchy and predominantly within
Fig 1. This representative cyostat section from Patient 3
reveals patchy, irregular reactivity f o r myosin-A TPase (PH
4.6) in many myojbers. Bar = 3 0 , ~ .
atrophic fibers (three specimens). Pan-myosin reactiv-
ity, assessed in 6 patients, correlated with the degree
of reactivity on myosin-ATPase staining in serial sec-
tions (Fig 2). Myofibers with reduced ATPase activity
had reduced myosin activity in a similar pattern; how-
ever, in Patient 13 the loss of myosin was more obvious
with myosin immunohistochemistry. Actin staining
was preserved within the areas of decreased myosin
staining. In many atrophic and some larger fibers with
and sometimes without irregular ATPase staining,
basophilic stippling was noted on H&E-stained sec-

Lacomis et al: Myopathy of Intensive Care 649


Fig 2. Serial cryostat sections from Patient 8. (A) Atrophic
angulated myojbers are present including two that did not
stain with ATPase at acid or alkaline pHs (arrows) (2TI’aj.e,
pH 3.4 is shown). (B) The atrophic myojbers stained hypoin-
tensely fir myosin. (C) Actin reactivity is preserved, and two
fibers have relatively intense reactivity (arrows). Bar = 25 ,u.
Fig 3. Electron micrograph from Patient 3 reveals loss of A
bands (A) and thick filaments with relative preservation of I
bands (frdmed by arrows) and thin filaments. Bar = 1.7 p.
tions, and NADH-TR-reacted fibers often had irregu-
lar staining consistent with a disruption in the inter-
myofibrillar network.
EM revealed selective disruption or loss of A bands
and thick filaments in some myocyres in 11 of the 11
specimens with irregular myosin-ATPase staining (Fig
3). Atrophic fibers were often more affected than nor-
mal-size fibers. Occasional myocytes had disruption of
all myofilaments. Linear aggregates of normal rnito-
chondria were often seen in nonnecrotic fibers. Dilata-
tion of the satcotubular system was focally present in
most specimens. The three biopsy specimens without
regions of absent ATPase staining did not have rela-
tively selective loss of thick filaments shown by EM.
Two of these three revealed nonspecific changes such
as Z-band streaming; the EM appearance was normal
in one. Intramuscular nerve twigs, present in frozen
tissue of 3 patients and in plastic-embedded sections
of 1, revealed no signs of axonal degeneration.
650 Annals of Neurology Vol 40 No 4 October 1996
Discussion
The incidence of AMIC is uncertain. We believe that
this retrospective study underestimates the incidence of
AMIC, especially milder forms, in our ICUs in which
there is an institutional referral bias toward organ
transplantation. At least in the status asthmaticus pop-
ulation, AMIC is probably very common when IV CSs
(of various types) and NMBAs are utilized. A prospec-
tive study of 25 patients treated for status asthmaticus
revealed that 9 (36%) developed myopathy-all re-
ceived both vecuronium and CSs-and 19 (76%) had
elevated CK levels [22].In that study, there was a posi-
tive correlation between the dose and duration of
NMBA therapy and the likelihood of developing my-
opathy. A positive correlation was not found regarding
CS doses and myopathy. However, most patients who
develop AMIC received at least the cumulative equiva-
lent dose of 1,000 mg of methylprednisolone. For ex-
ample, Shee [16] retrospectively noted that only the 4
of 9 patients with status asthmaticus who received
more than 5 gm of hydrocortisone (equivalent ro 1,000
mg of methylprednisolone) developed myopathy. In
our patients, the cumulative doses of methylpredniso-
lone (2,326 _f 1,036 mg, mean k standard deviation
[SD]) and the durations of treatment (12 days 2 6)
were within the ranges of those reported in most pa-
tients with AMIC. At our institution, most organ
transplant patients receive intraoperative NMBAs and
high-dose methylprednisolone (usually 1 gm followed
by a rapid taper to prevent graft rejection), placing
them at risk for AMIC. Only Patient 8 received less
than 1 gm of methylprednisolone (in addition to a low
dose of vecuronium), and she had mild weakness.
In our patients, the doses of NMBAs used were low
and the durations of administration short compared to
those for other reported patients with AMIC [ 16, 29-
311. Nevertheless, we did note that our patients who
received more than 80 mg of vecuronium in addition
to CSs were all quadriplegic (MRC 0) at the time of
diagnosis of AMIC, while most who received lower
doses of NMBAs were not as weak (MRC scores of
1-8). Due to the lack of a control group of patients
exposed to NMBAs and CSs who did not develop my-
opathy, we cannot determine whether the doses of
NMBAs were related to the development of myopathy.
Our ICU physicians generally now use “train of four”
monitoring and intermittent (rather than continuous)
NMBA administration to limit the total exposure to
NMBAs. Patient 10 was our only patient who did not
receive a NMBA. She had typical clinical, EMG, and
pathological features of AMIC, but she had only mild
weakness. Patient 14 received NMBAs during liver
transplantation, but he was not weak postoperatively.
Myopathy developed a week later only after he received
high-dose IV CSs (without NMBAs) for acute graft
rejection. He improved relatively quickly.
 As in other studies, vecuronium was the NMBA
used most commonly in our patients; 7 also received
pancuronium; 1 received doxacurium in addition to
vecuronium. Pancuronium and vecuronium are amino-
steroid compounds structurally related to CSs; how-
ever, others observed AMIC in association with non-
aminosteroids such as atracurium [28, 32, 331. It is
unknown whether the likelihood of developing AMIC
depends on the type of nondepolarizing NMBA uti-
lized.
In other reports, but not in our patient population,
a small percentage of critically ill patients who received
NMBAs (in very high doses) without CSs developed
clinical and EMG features of AMIC [30, 321. The
pathological findings included myofiber necrosis, but
loss of thick filaments was not observed. However,
muscle necrosis may be common in an ICU population
[41] for a number of reasons including sepsis and drug
exposures. In addition, there is neither an experimental
model of NMBA myopathy nor unique histopathology
in the reported patients. Therefore, it remains un-
proved whether prolonged NMBA treatment alone
(without CS exposure) causes the myopathy noted in
these rare patients.
Clinical and Laboratory Features
The neurological findings that we describe are typical
of AMIC. The precise onset of acute weakness is often
difficult to pinpoint due to NMBA and sedative ad-
ministration. Four of our patients were also initially
encephalopathic on a toxic-metabolic basis, further
masking their myopathic weakness. As we noted, there
is usually diffuse weakness often including respiratory
muscles. Our patients had uniform neck flexor weak-
ness, and we noted facial weakness more frequently
than in previous observations [16, 26, 29, 311; facial
weakness seemed to occur in the most severely affected
patients. O n the other hand, extraocular muscle
involvement is rare [16, 17, 29, 311, and we observed
it in only 1 patient. The weakness was reversible over
weeks to months if the underlying diseases were suc-
cessfully treated and the CSs and NMBAs were discon-
tinued.
Only a minority of our patients had an elevated CK
level despite the presence of pathological evidence of
muscle necrosis in the majority. In a review of AMIC
patients reported retrospectively, the CK level was ele-
vated in about half [21].We suspect that CK was usu-
ally measured too late (range, 9-41 days after exposure
to CSs) to detect an increase. In the prospective series
[22], the CK peak noted in patients with myopathy
occurred 3.6 ? 1.5 days after treatment, and the CK
was elevated for 9.8 2 5.9 days. These data suggest
that serial CK monitoring in patients at risk for AMIC
should be performed early after exposure to IV CSs
and NMBAs. A rising CK concentration could alert
the clinician to an evolving myopathy. If feasible, the
CS or NMBA, or both could be discontinued or given
at a lower dose.
In addition to CSs and NMBAs, metabolic factors
may play a role in potentiating AMIC. Infections, elec-
trolyte abnormalities, and hyperglycemia (presumably
due to CSs, sepsis, or preexisting diabetes mellitus)
were common in our patients. Hyperglycemia is also
common in patients who develop critical illness PN [2,
31. The majority of our patients also had disturbances
of renal or hepatic function. Patients with renal failure
[42] and perhaps liver failure [32] are more likely to
have prolongation of NMBA activity due to the persis-
tence of metabolites; thus, the likelihood of developing
NMBA-related neuromuscular toxicity may be in-
creased in these settings. Our liver transplant patients
who received vecuronium and pancuronium, which are
acetylated in the liver and partially excreted in bile (es-
pecially vecuronium) [43],may have been at particular
risk for NMBA toxicity during transplantation when
they temporarily lacked a functioning liver. The roles
of these metabolic factors as well as sepsis in potentiat-
ing AMIC are unknown and are difficult to study in
humans because these variables cannot be controlled.
Electrodiagnostic Findings
Electrodiagnostic studies are very useful in determining
the cause of flaccid weakness in ICU patients, espe-
cially when tendon reflexes are attenuated and the
sensory examination is unreliable. Given these signs,
which were present in most of our patients, the disease
process could be localized to any component of the
motor unit. In particular, we noted that AMIC was
very difficult to differentiate from critical illness PN by
clinical examination alone.
Our patients had rather typicaI eiectrodiagnostic
findings of AMIC (see Table 2). The presence of gener-
ally diffuse fibrillation potentials and short-duration
MUPs-sometimes with early recruitment-was com-
patible with a diffuse necrotizing myopathy. The low
motor amplitudes were presumably due to loss of myo-
fibers from necrosis. Purely on electrophysiological
grounds, a coexisting terminal motor axonopathy is dif-
ficult to exclude, but we and others [28, 291 were un-
able to histopathologically identify degeneration of mo-
tor axons in several patients. In addition, the reported
finding of electrical inexcitability of paralyzed muscles
in several patients with AMIC raises the possibility that
loss of sarcolemmai integrity alone or in addition to
necrosis accounts for the low motor amplitudes [44].
Occasionally, sensory amplitudes were mildly re-
duced (generally 1-2O'Yo below the lower limit of nor-
mal in one or two nerves), but the motor amplitude
reductions were substantially greater (generally <50%
of the lower limit of normal). Others believe that the
relatively minor reductions in sensory potentials in
Lacomis et al: Myopathy of Intensive Care 651
some patients with AMIC may be due to technical
factors such as edema at the recording site, as serial
studies show rather rapid amplitude improvement [31].
However, a coexisting mild neuropathic process could
still be present in some of these rare patients [21, 361
because the amplitude changes can also occur in non-
edematous regions. Given this rapidity of improvement
in amplitudes, the proposed neuropathic disturbance,
if present, is unlikely to stem from significant axonal
degeneration as is noted in critical illness PN (1-41.
Certainly, some of our patients who were exposed to
NMBAs and CSs were also septic and had multiorgan
failure, thus placing them at risk for critical illness PN
[l-41 as well as AMIC, but the relative sparing of sen-
sory amplitudes in the setting of low motoi- responses
and the early recruitment of short-duration MUPs
were the electrodiagnostic hallmarks that distinguished
AMIC from critical illness PN.
In contrast, 2 of our patients did have sensory ampli-
tude reductions that were as severe as the motor abnor-
malities. Patient 14 had reduced upper-extremity sen-
sory responses and no sural response, but the sural
study was unreliable due to edema. Conduction veloci-
ties were normal. Patient 12, who had a history of
alcoholism, had reduced but present sensory (including
sural) and motor responses as well as mild slowing of
conduction velocities. We also noted fiber-type group-
ing (consistent with reinnervation) in his muscle biopsy
specimen, obtained only 10 days after liver transplanta-
tion, suggestive of an underlying chronic alcoholic
polyneuropathy. In both patients, the neurological ex-
amination revealed greater proximal than distal weak-
ness with normal sensation. Myopathic changes were
also present in the biopsy specimen from Patient 14.
Although an element of critical illness I”cannot be
excluded in these patients, the distribution of weakness,
rate of recovery, lack of both sepsis and multiorgan
failure initially, and the EMG and histopathological
findings (Patient 14) are most consistent with AMIC
being the cause of acute weakness.
We did not identify prolonged NMJ blockade as a
cause of weakness in the patients who were exposed to
NMBAs despite the presence of renal [42] or hepatic
failure [32]. Barohn and colleagues [32] showed that
AMIC is preceded by NMJ blockade in some patients.
It is possible that prolonged NMJ blockade occurred
earlier in some of our patients prior to performance of
the EMG, or that it was too subtle to be detected by
repetitive stimulation alone. Nevertheless, their persis-
tent weakness was due to myopathy.
Muscle Histopathology
In this relatively large series of AMIC patients whose
evaluations included muscle biopsy, we found myo-
pathic changes with selective thick (myosin) filament
loss in the majority. In addition, myofiber necrosis was
652 Annals of Neurology Vol 40 No 4 October 1996
noted in other regions of these specimens. Patchy re-
ductions in enzyme activity or absent staining of myo-
fibers on serial ATPase-reacted sections at acid and al-
kaline pHs were predictive of thick filament loss that
was confirmed by EM. We found that myosin heavy-
chain and myosin-ATPase reactivity correlated. Actin
was present in regions of decreased myosin staining.
Because patchy loss of ATPase reactivity could occur
due to early myofiber necrosis, which should affect all
myofilaments, it is iniportant to determine that there
is preservation of actin (thin filaments), either immu-
nohistochemically or by ultrastructural evaluation.
The loss of thick filaments was generally widespread
but patchy within affected fibers of varying sizes and
of both fiber types. Atrophic fibers were often most
affected. Type 2 fibers, those most vulnerable to CS
and disuse atrophy, were particularly involved. Others
noted even more selective involvement of type 2 fibers
in 2 patients with critical illness and myopathic features
[45]. In Patient 8, myosin loss occurred only within
the atrophic fibers in a diffuse rather than patchy pat-
tern, which we consider to be part of the spectrum of
AMIC [46]rather than a unique finding [47].Atrophic
fibers may be affected earlier in the course of AMIC,
but serial muscle biopsies in a dose-escalation model
would be required to determine whether the patterns
of myosin loss change with dosing and with time. In
the animal model, the degree of thick filament loss
does increase during the weeks after CS exposure [37].
Loss of A-band density is evident by 7 days but only
in a few fibers. At 13 days, many fibers are affected,
and almost all are affected at 28 days; however, there
is variability in thick filament loss among different re-
gions of the same muscle as shown by EM [37].
In 3 patients, muscle biopsy specimens did not re-
veal thick filament loss. These biopsies were performed
earlier (12 -+ 2 days, mean 2 SD) than those that did
reveal thick filament loss (30 5 11, p < 0.02). In
Patient 13, the biopsy was also performed relatively
early, and the myosin loss was subtle. It is likely that
the appearance of thick filament loss lags behind the
development of clinical weakness, as suggested by the
animal model. It is less likely that the more subtle loss
of thick filaments that might occur “early” was missed
due to sampling error, given the relatively diffuse distri-
bution of weakness and EMG abnormalities. Alterna-
tively, the acute myopathy without thick filament loss
could have a different pathogenesis. We were reluctant
to evaluate these possibilities further by performing
multiple or serial muscle biopsies on our patients.
Pathogenesis of Loss of Thick Filaments
Loss of thick filaments is not specific. It also occurs
sporadically in the setting of dermatomyositis, throm-
botic thrombocytopenia purpura, human immunode-
ficiency virus infection, and congenital myopathy [48,
Corticosteroids
+
=/\t=
NMBAsor
-
- I Band-
ABand-
NMBAs
Renal, Hepatic Failure
Abnormal NMJ; 7 Metabolic Factors of
y y
Denervation; Disuse
4 \ Critical Illness
i to control in humans but could be examined in animal
Fig 4. A theoretical model of acute myopathy of intensive
care. Loss of A bands and thick filaments occurs ajer expo-
sure to high-dose intravenous corticosteroids, but neuromuscu-
lar junction-blocking agents (NMBAs), denervation, other
causes of a motor end-plate disturbance, or d i k e are neces-
say to tvigger the process. Overt myofiber necrosis (with disor-
ganization of all myofihments) can alro result from this
combination of fdctors. NMBAs and metabolic disturbances
associated with critical illness may also induce a necrotizing
myopathy without selective loss of thick jlaments. NMBA
“toxicity,” including prolonged neuromuscular junction
(NMJ) blockade, may be intensified in the setting of renal or
hepatic failure. Z = Z line.
491. Sher and colleagues [50] first reported an acute
myopathy with “selective lysis of myosin filaments” in
the ICU setting in 1979 in a woman with allergic vas-
culitis, renal failure, and pneumonia. She received IV
CSs; NMBA use was not noted. She recovered in 6
weeks. Danon and Carpenter [18] first described loss
of thick filaments in AMIC associated with NMBAs
and IV CSs and other reports followed [20, 25, 29,
321. Despite the lack of specificity, we agree with Da-
non and Carpenter that a loss of thick filaments noted
in acute myopathy associated with IV CS and NMBAs
denotes a distinctive syndrome.
The precise cause of the loss of thick filaments in
AMIC is uncertain. In the animal model, disassembly
of myosin monomers-presumably due to an effect at
the level of cellular protein regulation-is thought to
lead to the loss of thick filaments [37,381. The experi-
mental animal model, our study, and the work of oth-
ers support the notion that IV CSs cause the loss of
thick filaments, but that other factors may “trigger”
the process [29] (Fig 4). An abnormal NMJ from phar-
macological blockade in ICU patients and an abnormal
end-plate following traumatic denervation in the ani-
mal model are likely triggers. Our study showed that
the NMBA exposure may be minimal. Also in support
of the view that an abnormal motor end-plate can in-
cite this process is the fact that a patient with a dys-
functional end-plate from myasthenia gravis developed
loss of thick filaments after receiving high-dose CSs
[51]. Disuse (or immobilization) that leads to an in-
crease in CS receptors [52] may also lower the thresh-
old for this toxic myopathy. Interestingly, Achilles te-
notomy in rats, which leads to soleus shortening and
immobilization, causes corelike lesions with some pref-
erential loss of thick filaments [53].Other factors that
may coexist in AMIC patients and alter the nerve and
motor end-plate milieu, including critical illness I“,
sepsis, cytokines, and hyperglycemia, may also play a
role in potentiating AMIC [54](see Fig 4 ) . These vari-
ables require further study; however, they are difficult
models.
W e appreciate the excellent assistance from our neuromuscular fel-
lows and histopathology and neurodiagnostic technicians. Clayton
Wiley, MD, PhD, kindly reviewed the manuscript and offered
thoughtful suggestions. Regina Hardison provided statistical guid-
ance.
References
1. Bolton CF, Gilbert JJ, Hahn AF, Sibbald WJ. Polyneuropathy
in critically ill patients. J Neurol Neurosurg Psychiatry 1984;
47: 1223- 1231
2. Zochodne D W , Bolron CF, Wells GA, et al. Critical illness
polyneuropathy. A complication of sepsis and multiple organ
failure. Brain 1987;110:819-842
3. Witt NJ, Zochodne D W , Bolton CF, et al. Peripheral nerve
function in sepsis and multiorgan failure. Chest 1991;99: 176-
184
4. Bolton CF, Young GB, Zochodne DW. The neurological com-
plications of sepsis. Ann Neurol 1993;33:94-100
5. O p de Coul AAW, Lambregts PC, Koeman J, et al. Neuromus-
cular complications in parients given Pavulon (pancuronium
bromide) during artificial ventilation. Clin Neurol Neurosurg
1985;87: 17-22
6. Segredo V, Marthay MA, Sharma ML, et al. Prolonged neuro-
muscular blockade after long-term administration of vecuro-
nium in two critically ill patients. Anesthesiology 1990;72:
566-570
7. Partridge BL, Abroms J H , Basemore C , Rubin R. Prolonged
neuromuscular blockade after long-term infusion of vecuro-
nium bromide in the intensive care unit. Crit Care Med 1990;
18:1177-1179
8. MacFarlane IA, Rosenthal FD. Severe myopathy after status
asthmaticus. Lancet 1977;2:615
9. Van Mark W , Woods KL. Acute hydrocortisone myopathy.
BMJ 1980;281:271-272
10. Knox AJ, Mascie-Taylor BH, Muers MF. Acute hydrocortisone
myopathy in acute severe asthma. Thorax 1986;41:411-412
11. Kaplan PW, Rocha W , Sanders DB, et al. Acute steroid-
induced tetraplegia following status asthmaticus. Pediatrics
1986;78: 121-123
12. Bachmann P, Gassorgues P, Piperno D, et al. Myopathie aigue
au decors de l’etat de ma1 asthmatique. Presse Med 1987;16:
1486
13. Williams TJ, O’Heir RE, Czarny D, et al. Acute myopathy in
severe asthma treated with intravenously administered cortico-
steroids. Am Rev Respir Dis 1988;137:460-463
14. Sury MRJ, Russell G N , Heaf DI’. Hydrocortisone myopathy.
Lancet 1988;2:515 (Letter)
15. Brun-Buisson C, Gherardi R. Hydrocortisone and pancuro-
Lacomis et al: Myopathy of Intensive Care 653
 nium bromide. Acute myopathy during status asthmaticus. Crit
Care Med 1988;16:731 (Letter)
16. Shee CD. Risk factors for hydrocortisone myopathy in acute
severe asthma. Respir Med 1990;84:229-233
17. Sitwell LD, Weinshenker BG, Monpetit V, Reid D. Complete
ophthalmoplegia as a complication of acute corticosreroid- and
pancuronium-associated myoparhy. Neurology 1991;41:921-
922
18. Danon MJ, Carpenter S. Myopathy with thick filament (myo-
sin) loss following prolonged paralysis with vecuronium during
steroid treatment. Muscle Nerve 1991;14:1131-1139
19. De Smet Y, Jaminet M, Jaeger U, et al. Myopathie aigue corti-
sonique de I’asthmatique. Rev Neurol (Paris) 1991;147:682-
685
20. Hitano M, Ott BR, Raps EC, er al. Acute quadriplegic myopa-
thy: a complication o f treatment with steroids, nondepolarizing
blocking agents, or both. Neurology 1992;42:2082-2087
21. Lacomis D, Smith TW, Chad DA. Acute myopathy and neu-
ropathy in status asthmaticus: case report and literature review.
Muscle Nerve 1993;16:84-90
22. Douglas JA, Tuxen DV, Horne M, et al. Myopathy in severe
asthma. Am Rev Respir Dis 1992;146:5 17-5 19
23. Griffin D, Fairrnan N, Coursin D, et al. Acute myopathy dur-
ing treatment of status asthmaticus with corticosteroids and
steroidal muscle relaxants. Chest 1992;102:510-514
24. Bird SJ, Mackin GA, Schotland DL, Raps EC. Acute rnyo-
parhic quadriplegia: a unique syndrome associated with vecuro-
nium and steroid treatment. Muscle Nerve 1332;15:1208 (Ab-
stract)
25. Waclawik AJ, Sufit RL, Beinlich BR, Schutta HS. Acute niyop-
athy with selective degeneration of myosin thick filaments fol-
lowing status asthmaticus treated with methylprednisolone and
vecuronium. Neuromuscul Disord 19922: 19-26
26. Blackie JD, Gibson P, Murree-Allen K, Saul WP. Acute myop-
athy in status asthmaticus. Clin Exp Neurol 1993;30:72-81
27. Mackie G, Road J, Stewart H, Eisen A. Reversible paraly-
sis with status asthmaticus, steroids and pancuronium: clini-
cal electrophysiological correlates. Neurology 1993;43:A166-
A I67 (Abstract)
28. Bella IR, Chad DA, Smith TW,et al. Ophrhalmoplegia and
quadriplegia in the wake of intensive therapy for status asth-
maricus. Muscle Nerve 1994;17:1122- 1123 (Abstract)
29. Ramsay DA, Zochodne DW, Robertson DM, et al. A syn-
drome of acute severe muscle necrosis in intensive care unit
patients. J Neuroparhol Exp Neurol 1993;52:387-398
30. Gooch JL, Moore MH, Ryser DK. Prolonged paralysis after
neuromuscular junction blockade: case reports and electrodiag-
nosric findings. Arch Phys Med Rehabil 1993;74:1007-1011
31. Zochodne DW, Ramsay DA, Saly V, et al. Acute necrotizing
myoparhy o f intensive care: electrophysiologicalstudies. Muscle
Nerve 1394;17:285-292
32. Barohn RJ, Jackson CE, Rogers SJ, et al. Prolonged paralysis
due to nondepolarizing neuromuscular blocking agents and
corticosteroids. Muscle Nerve 1994;17:647-654
33. Miro 0 , Grau JM, Nadal P, et al. Acute myopathy related
to the administration of ghcocorticoids and neuromuscular
blockers. Med Clin (Barc) 1994;103:458-460
34. Gooch JL. AAEM case report #29. Prolonged paralysis after
neuromuscular blockade. Muscle Nerve 1995;18:937-945
654 Annals of Neurology Vol 40 No 4 October 1996
35. Subramony SH, Carpenter DE, Raju S, et al. Myopathy and
prolonged neuromuscular blockade after lung transplant. Crit
Care Med 1991;19:1580-1582
36. Faragher MW, Day BJ, Dennett X. Critical care myopathy: an
electrophysiologicaland histological study. Muscle Nerve 1996;
19:516-5 18
37. Rouleau G, Karpati G, Carpenter S, et al. Glucocorticoid ex-
cess induces preferential depletion of myosin in denervared
skeletal muscle fibers. Muscle Nerve 1987;428-438
38. Massa R, Carpenter S, Holland P, Karpati G. Loss and renewal
of thick filaments in glucocorricoid-treated rat soleus after de-
nervation and reinnewation. Muscle Nerve 1992;15:1290-
1298
39. Brooke MH, Fenichel GM, Griggs RC, et al. Clinical invesri-
gation in Duchenne dystrophy: 2. Determination of the
“power” of therapeutic trials based on the natural history.
Muscle Nerve 1983;6:91-103
40. Cot; AM, Jimknez L, Adelman LS, Munsat TL. Needle muscle
biopsy with the automatic Biopty instrument. Neurology 1992;
42:2212-2213
41. Helliwell TR, Coakley JH, Wagenmakers AJM, et al. Necrotiz-
ing myopathy in critically-ill patients. J Pathol 1991;164:307-
314
42. Segredo V, Caldwell JE, Matthay MA, et al. Persistent paralysis
in critically ill patients after long-term administration of vecu-
ronium. N Engl J Med 1992;327:524-528
43. Hunter JM. New neuromuscular blocking drugs. N Engl J
Med 1995;332: 1691-1699
44. Rich MM, Teener JW, Raps EC, et al. Muscle is electrically
inexcitable in acute quadriplegic myoparhy. Neurology 1996;
46731-736
45. Gutrnann L, Blumenthal D, Gutman L, Schochet SS. Acute
type I1 myofiber atrophy in critical illness. Neurology 1996;
46:819-821
46. Chad DA, Lacomis D. Critically ill patients with newly ac-
quired weakness: the cliiiicopathological spectrum. Ann Neurol
1994;35:257-259 (Editorial)
47. Al-Lozzi M, Pestronk A, Yee WC, et al. Rapidly evolving my-
opathy with myosin-deficient muscle fibers. Ann Neurol 1994;
35~273-279
48. Carpenter S, Karpati G. Pathology of skeletal muscle. New
York: Churchill Livingstone, 1984:220-221
49. Gonzales MF, Olney RK, So YT, et al. Subacute structural
myopathy associated with human immunodeficiency virus in-
fection. Arch Neurol 1988;45:585-587
50. Sher JH, Shafiq SA, Schutta HS. Acute myopathy with selec-
tive lysis of myosin filaments. Neurology 1979;29:100-
106
51. Panegyres PK, Squier M, Mills KR, Newsom-Davis J. Acute
niyopathy associated with large parenteral doses of corticoste-
roid in myasthenia gravis. J Neurol Neurosurg Psychiatry 1993;
56:702-704
52. DuBois DC, Almon RR. A possible role for glucocorticoids in
denervation atrophy. Muscle Nerve 1991;14:1131-1139
53. Kai-pati G, Carpenter S, Eisen A. Experimental core-like lesions
and nemaline rods. A correlative morphological and physiologi-
cal study. Arch Neurol 1972;27:237-25 1
54. Bolton CF. Neuromuscular complications of sepsis. Intensive
Care Med 1993;19:S58-S63