Download Regional Ischemic Immune Myopathy: A Paraneoplastic

J Neuropathol Exp Neurol
Copyright Ó 2014 by the American Association of Neuropathologists, Inc.
Vol. 73, No. 12
December 2014
pp. 1126Y1133
ORIGINAL ARTICLE
Regional Ischemic Immune Myopathy: A Paraneoplastic
Dermatomyopathy
Chunyu Cai, MD, PhD, Ali Alshehri, MD, Rati Choksi, MS, and Alan Pestronk, MD
Abstract
Necrosis and regeneration of scattered muscle fibers are common
features of many active acquired and immune myopathies. We studied
a series of patients with acquired myopathies with an unusual pattern of
regional, rather than scattered, muscle fiber necrosis and regeneration.
Retrospective review of records of 7 patients with acquired myopathies
having regional muscle fiber necrosis on muscle biopsy. Clinical features of patients included proximal symmetric weakness in arms and
legs with a subacute onset (100%) beginning at ages between 41 and
92 years, with dysphagia (83%), myalgias (100%), skin rash (67%),
and associated malignancy (71%). Serum creatine kinase was often
very high (91,600 U/L) (83%). Survival was less than 1 year in 43%.
Myopathology included a regional distribution with muscle fiber
necrosis and capillary loss in the border zones between intermediatesized perimysial vessels, vascular pathology with damaged walls of
intermediate-sized perimysial veins, and connective tissue with expression of the ischemia marker carbonic anhydrase IX but no mononuclear
cell inflammatory foci. These data indicate that regional ischemic immune myopathies are likely caused by ischemia in border zones between damaged intermediate-sized perimysial blood vessels. Regional
ischemic immune myopathies are a distinctive pathologic group of acquired, probably immune, noninflammatory dermatomyopathies with
weakness and often a skin rash and systemic neoplasm.
Key Words: Dermatomyositis, Immune, Ischemia, Myopathy, Vein.
tissue. Immune polymyopathies are active myopathies with
necrosis and regeneration of scattered muscle fibers and increased endomysial connective tissue but little mononuclear
cell inflammation. Histiocytic inflammatory myopathies have
foci of histiocytic cells and focal invasion of muscle fibers by
cells. Inflammatory myopathies with vacuoles, aggregates, and
mitochondrial pathology (IM-VAMP) have morphologic abnormalities in myofibers and endomysial mononuclear cell inflammatory infiltrates. IM-VAMP including inclusion body
myositis as a subgroup, are slowly progressive myopathies with
patchy weakness and poor response to treatment. Although
further documentation is necessary, tissue pathology in muscle,
such as vessel or connective tissue damage, may also occur in
other organs and explain myopathy syndromes with systemic
features such as skin rash or lung disease.
We studied features of patients with an acquired myopathy with an unusual pattern of regional, rather than scattered,
muscle fiber necrosis and regeneration and little inflammation.
Each individual region of myopathology contained clusters of
damaged muscle fibers in a similar stage of necrosis or regeneration. We analyzed the patterns of muscle fiber and vascular
and connective tissue pathology in these patients. Our results
suggest that this pathologic subgroup of IIM is likely to be a
myovasculopathy. Common clinical features in patients with
this regional, probably ischemic and immune, dermatomyopathy
(RIIM) are proximal symmetric weakness, a skin rash, associated neoplasms, and a high serum creatine kinase (CK).
INTRODUCTION
Myopathologic features can be used to characterize and
classify acquired immune and inflammatory myopathy (IIM)
syndromes. Several IIM categories can be defined based on
patterns of immune features and types of pathology in muscle
fibers, connective tissues, and vessels (1). Immune myopathies
with perimysial pathology (IMPP) have connective tissue
damage in the perimysium with fragmentation and histiocytic
cellularity. Myovasculopathies have abnormal vessels in the
perimysium or endomysium that may produce muscle ischemia.
Immune myopathies with endomysial pathology often have
diffuse C5b-9 complement deposition on endomysial connective
From the Departments of Pathology and Immunology (CC, AP), and Neurology (AA, RC, AP), Washington University School of Medicine, St.
Louis, Missouri.
Send correspondence and reprint requests to: Alan Pestronk, MD, Department of
Neurology, Washington University School of Medicine, 660 South Euclid
Ave, Box 8111, St. Louis, MO 63110; E-mail: [email protected]
The authors report no conflicts of interest.
1126
MATERIALS AND METHODS
Patient Selection Criteria
We retrospectively reviewed the muscle biopsy database from the neuromuscular service at Washington University School of Medicine in St. Louis between 1996 and 2013.
We identified 7 adults who had muscle biopsies with an unusual multiregional pattern of myopathology, with areas of at
least 5 contiguous muscle fibers in similar stages of necrosis
or regeneration (Table 1). This myopathology pattern differs
from most previously described immune ‘‘necrotic’’ myopathies that have individual muscle fibers in varying stages of
necrosis and regeneration scattered through the biopsy rather
than contiguous fibers with a uniform stage of damage. Clinical
and laboratory data were obtained from patient clinical records
and biopsy reports. Five patients (Patients 2Y6) were evaluated
and treated at the Neuromuscular Service at Washington University School of Medicine and 2 (Patients 1 and 7) had clinical
J Neuropathol Exp Neurol Volume 73, Number 12, December 2014
Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.
J Neuropathol Exp Neurol Volume 73, Number 12, December 2014
Regional Ischemic Myopathy
TABLE 1. Clinical and Laboratory Characteristics of the 7 Patients
Patient No.
Demographics
Sex
Onset age, years
Clinical features
Weakness
Proximal 9 distal
Maximal severity, weeks
Dysphagia
Skin rash
Myalgia
Neoplasm association
Steroid treatment response
Follow-up, months
Laboratory data
Serum creatine kinase (maximum), IU/L
Serum aldolase (maximum), IU/L
Serum anti-nuclear antibody titer
EMG (irritable myopathy)
1
2
3
4
5
6
7
M
77
F
92
F
63
F
79
F
70
M
41
F
58
+
+
NA
NA
NA
+
Stomach; prostate
NA
4 Death
+
NA
1
+
+
+
Endometrial
Poor
1 Death
+
+
2
+
+
+
Lung, small cell
Yes
36 Death
+
+
4
+
+
+
Renal
Partial
18
+
+
2
j
j
+
Ovary; myeloma
Yes
4 Death
+
+
2
+
+
+
None
Partial
84
+
NA
6
+
j
NA
None
NA
12
12,420
NA
NA
NA
3,784
NA
0
NA
1,604
NA
1:320
+
4,162
17.2
1:1280
+
145
3.9
0
NA
217,755
898
1:1280
+
NA
NA
NA
NA
+, present; j, absent; ANA, anti-nuclear antibody; EMG, electromyography; F, female; M, male; NA, no data available.
data and muscle biopsies referred from outside institutions
(Table 1). Patient 1 has been previously reported (2). The
median time between the onset of weakness and muscle biopsy
was 5 weeks (range, 2.5Y13 weeks). Frozen muscle blocks and
muscle histochemistry slides were available for all patients.
Paraffin- or plastic-embedded blocks were available for 3 patients
(Patients 2, 4, and 5). All slides were independently reviewed
and evaluated by 2 muscle pathologists. As controls for carbonic
anhydrase IX (CA9) staining, we evaluated 4 normal muscles
and a series of 45 consecutive muscle biopsies performed for
evaluation of possible immune myopathies. The human studies
committee of Washington University in St. Louis approved
all procedures.
immersion in 3% glutaraldehyde in sodium phosphate buffer,
pH 7.4, overnight at 4-C. Tissue samples were postfixed in
phosphate buffered 2% OsO4, dehydrated in graded concentrations of ethanol, and embedded in Embed-812 (Electron
Microscopy Sciences, Hatfield, PA), with propylene oxide as an
intermediary solvent. One-micrometer-thick plastic sections
were examined by light microscopy after staining with toluidine
blue. Ultrathin sections of muscle biopsies were cut into mesh
or Formvar-coated slot grids, which permit visualization of entire fascicular cross sections. Tissues were subsequently stained
with uranyl acetate and lead citrate and examined using a JEOL
1200 electron microscope with an AMT digital camera.
Histochemistry and Immunohistochemistry
Cryostat sections of rapidly frozen muscle were processed
for muscle histochemistry, as previously described (3, 4). Immunohistochemical stains for antibodies were performed on
consecutive sections using muscle from patients and paired
controls on the same glass slide. Primary antibodies were directed
against CD68, CD3, CD4, and C5b-9 complement (membrane
attack complex) (Sigma-Aldrich, St. Louis, MO); class I human
major histocompatibility complex ([MHC-1] US Biological,
Swampscott, MA); platelet-endothelial cell adhesion molecule
1 ([PECAM-1, CD31] Lifespan Biosciences, Seattle, WA);
smooth muscle actin (Dako, Carpinteria, CA); collagen IV
(Chemicon, Temecula, CA); and CA9 (2D3, Novus, Littleton,
CO). Secondary antibodies were conjugated to green or red
fluorescent markers (anti-mouse Alexa Fluor 488; anti-rabbit
Alexa Fluor 594, Invitrogen, Carlsbad, CA). Ulex Europaeus
Agglutinin I lectin (Sigma-Aldrich) was used to visualize vascular endothelial cells.
Electron Microscopy
Tissues from 3 patients were available for ultrastructural analysis. A portion of each muscle biopsy was fixed by
RESULTS
Case Report
A 79-year-old white woman (Patient 4) presented with a
1-month history of skin rash on her arms and trunk, progressive weakness of her arms and legs, dysphagia, and muscle
discomfort. Two weeks after onset, she was unable to climb
stairs and had difficulty lifting her arms above her head to
brush her hair. Swallowing and chewing were impaired. She
had had a renal cell carcinoma and nephrectomy 5 years earlier. Losartan was started near the time of symptom onset.
Examination demonstrated moderately severe, symmetric, proximal predominant weakness in the arms and legs with normal
sensation. Serum CK was 4,133 IU/L. Electromyography showed
an irritable myopathy. Muscle pathology showed several regions
of muscle fiber necrosis. Investigation for associated malignancy
showed several lung nodules. Treatment with corticosteroids was
followed by resolution of the rash, no improvement in strength
or dysphagia, and a reduction in serum CK. She developed
several episodes of aspiration pneumonia and had gastric tube
placement 9 months after her initial presentation. Follow-up
chest computed tomography and positron emission tomography
Ó 2014 American Association of Neuropathologists, Inc.
Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.
1127
J Neuropathol Exp Neurol Volume 73, Number 12, December 2014
Cai et al
scans demonstrated many bilateral pulmonary nodules compatible
with progressive metastatic disease. Death occurred 18 months
after disease onset.
Clinical Features
Five women and two men developed weakness with onset at ages from 41 to 92 years (Table 1). All had a subacute
onset of proximal-greater-than-distal symmetric weakness that
progressed to severe disability across several days to 6 weeks.
Symptoms and signs included myalgias (100%), dysphagia
(83%), skin rash (67%), and limb swelling (28%). The skin
rash was located primarily on the face and chest, in some also
on dorsal arms and hands, and preceded weakness in some
patients (50%). No specific dermatologic features such as Gottron
papules or mechanic’s hands were described. Systemic pulmonary, cardiac, or gastrointestinal involvement was not present in
any patient. Five patients had associated malignancies (Patients
1Y5): 2 were newly discovered malignancies within weeks after
the onset of weakness (Patients 2 and 3); 2 patients had metastatic disease with known malignancies (Patients 4 and 5); and
1 patient had known neoplasms at the time of the onset of
the myopathy (Patient 1). The 5 patients with malignancies had
myopathy onset at ages from 63 to 92 years. The 2 patients with
no malignancies had onset at ages of 41 and 58 years. Serum
CK was often high (83%). Electromyography in 2 patients
showed an irritable myopathy with small motor units and fibrillations. Response to steroid treatment included marked improvement in strength beginning days after initiation in 2 patients
and minimal response in 2 others. Four of the 5 patients with
malignancy died within 3 years. One patient without malignancy
was alive and well after 7 years.
Muscle Pathology
Muscle pathology was regional (Table 2). Distinctive regions included necrotic muscle fiber clusters, zones of perimysium with abnormal intermediate-sized vessels (86%), purlieu
TABLE 2. Muscle Pathology in Patients With Regional Necrotic Myopathy
Patient No.
Necrotic zone
Muscle fibers
Necrosis, clustered*
Hypercontraction
Regeneration
Invasion by cells
MHC-1 upregulation
Capillaries
Loss
C5b9 deposition
Endomysium
Mononuclear cells
Acid phosphataseYpositive cells
Alkaline phosphatase positive
Edema
CA9
Perimysial veins
Wall structure abnormal†
Cells in wall
Surrounding perimysium damaged‡
Fibrinoid necrosis
Purlieu zone
Muscle fibers
Smallness
Immaturity (2C fibers)
MHC-1 upregulation
Capillaries
Enlargement
Alkaline phosphatase positive
C5b9 deposition
Connective tissue
Acid phosphataseYpositive cells
Alkaline phosphatase positive
1
2
3
4
5
6
7
+
+
j
j
j
+
+
j
j
j
+
+
j
j
j
+
+
j
j
j
+
+
j
j
j
+
+
j
j
+
+
+
j
j
j
+
j
T
j
+
j
+
j
+
j
j
j
+
j
j
T
j
+
+
j
T
j
+
+
j
j
j
+
+
j
+
j
+
+
j
+
j
+
+
j
+
+
+
j
j
+
j
+
+
+
+
+
j
+
+
+
j
+
+
+
j
+
+
+
j
+
+
+
j
j
j
j
j
+
+
+
j
+
+
j
+
+
+
+
+
j
+
+
+
+
+
+
j
j
j
+
+
+
+
+
j
+
+
+
+
+
+
+
+
+
+
+
+
j
j
j
+
+
+
+
+
+
+
+
+
+
+
+
+
j
j
+
+
*Muscle fiber groups with cytoplasm staining pale on H&E and NADH and strong for C5b-9 complement.
†Disrupted external elastic lamina or patchy loss of smooth muscle in the wall.
‡Fragmented surrounding perimysial connective tissue with acid phosphataseYpositive histiocytes and alkaline phosphatase staining.
+, feature present; T, minimal; j, absent; MHC-1, class I human major histocompatibility complex.
1128
Ó 2014 American Association of Neuropathologists, Inc.
Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.
J Neuropathol Exp Neurol Volume 73, Number 12, December 2014
FIGURE 1. Pattern of damage to muscle fibers and other structures
in regional ischemic immune myopathies (RIIM). (A) A cluster of
necrotic and some hypercontracted muscle fibers (right) separated by endomysial edema. Muscle fibers in the purlieu between
the artery-vein pair and the region of necrosis are viable (H&E
stain). (B) Regional distribution of muscle fiber necrosis in RIIM.
Anti-C5b-9 (green) stains the cytoplasm of necrotic muscle fibers
that are present in border zone regions between vessels.
AntiYcollagen IV (red) stains the basement membrane of all
muscle fibers and vessels. The arrowhead points to intermediatesized perimysial vessels located at the center of a vascular territory; these are the same vessels as in A. (C) Carbonic anhydrase
IX, an ischemia marker, is strongly expressed in the connective
tissue (perimysium and endomysium) and myofibers in the
border zone region with necrotic muscle fibers (serial section of
B). The arrowhead points to intermediate-sized perimysial vessels located at the center of a vascular territory (same vessels as
in A and B). (D) Border zone pathology. Necrotic muscle fibers
are located at the periphery of vascular territories rather than
along the perimysium. This pattern of muscle fiber pathology is
different from the selective atrophy of muscle fibers at the edges
of fascicles seen in childhood dermatomyositis with perivascular
inflammation. Myofibers near the perimysial vessels (arrowheads)
are viable (C5b-9 (green) and collagen IV (red) double labeling).
Panels (AYC) are from Patient 7, panel (D) is from Patient 4.
Scale bars = (A) 100 Km; (BYD) 200 Km.
Regional Ischemic Myopathy
on H&E, and dark staining for COX and C5b-9. Phagocytic
infiltration of necrotic fibers was minimal except occasionally
near the periphery of the clusters. Class I human major histocompatibility complex staining of muscle fibers was mildly increased. There was apparent edema with clear spaces between
muscle fibers. Numbers of endomysial capillaries were markedly
reduced. The few remaining capillary remnants showed collagen4 staining, no PECAM-stained endothelium, and usually had
C5b-9 deposition and alkaline phosphatase staining. Carbonic
anhydrase IX, a hypoxia-inducible marker (5, 6), was strongly
expressed in the necrotic zone, with intense staining on the perimysium (83%), endomysium (100%), and, with less intensity, in
the cytoplasm of necrotic myofibers (83%) (Fig. 1C). In normal
control muscles, CA9 was present in endomysial capillaries. In
45 consecutive disease control biopsies evaluated for immune
disorders, CA9 was sometimes present in inflammatory cells,
particularly in biopsies with abundant acid phosphataseYpositive
cells. Carbonic anhydrase IX showed occasional intense staining
in the perimysium (16%) but rarely in the endomysium (2%) or
necrotic muscle fiber cytoplasm (2%).
Perimysial Intermediate-Sized Vessels
Perimysial intermediate-sized vessels (Fig. 1, arrowheads)
were generally separated from areas of muscle fiber necrosis by
a ring or zone of viable myofibers, the purlieu. Structural damage to intermediate-sized perimysial vessels was common, with
veins typically showing more pathology than arteries. Abnormalities of veins included thickened walls (Fig. 2B, C), intramural and perivascular scattered inflammatory cells (often acid
(i.e. surrounding) areas with enlarged capillaries and muscle fibers with irregular internal architecture that were located between
perimysial vessels and necrotic fiber clusters (86%), and groups of
small regenerating muscle fibers (43%). In most patients (86%),
the clusters of muscle fiber necrosis were clearly localized to
border zones in territories between intermediate-sized vessels.
Necrotic Muscle Fiber Clusters
Necrotic muscle fibers were contiguous and in similar
stages of early pathology. Hematoxylin and eosin (H&E) stains
showed many neighboring muscle fibers with featureless pink or
pale-stained cytoplasm (Fig. 1A). Necrotic muscle fiber cytoplasm
had reduced or absent staining for NADH, cytochrome oxidase
(COX), and succinate dehydrogenase but strong diffuse staining
for C5b-9 (Fig. 1B, D), IgM, and C4d. Some myofibers were
hypercontracted with rounded contour, eosinophilic cytoplasm
FIGURE 2. Vascular damage is more prominent in the perimysial
veins than in arteries. (A) Normal vein and artery from a patient
with no myopathy (Verhoeff van Gieson [VvG] stain). (B, C) Veins
have swollen endothelial cells, thickened walls, intramural inflammatory cells, and damaged elastic lamina. The arteries are normal
(VvG). (D) Patchy loss of smooth muscle in the vein wall (arrowheads). The artery appears normal (double labeling for platelet
endothelial cell adhesion molecule [PECAM, red] and smooth
muscle actin [green]). a, artery; v, vein. Scale bar = (AYD) 50 Km.
Ó 2014 American Association of Neuropathologists, Inc.
Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.
1129
J Neuropathol Exp Neurol Volume 73, Number 12, December 2014
Cai et al
phosphatase positive) (Fig. 2B, C), disrupted external elastic
lamina (Fig. 2C), and patchy loss of smooth muscle in the wall
(Fig. 2D). Perimysial connective tissue surrounding the vessels
was often fragmented, contained acid phosphataseYpositive
histiocytes, and stained strongly for alkaline phosphatase. The
morphology of the neighboring perimysial arteries was generally unremarkable (Fig. 2BYD).
Purlieu Areas
Purlieu areas extended between necrotic muscle fiber clusters and abnormal intermediate-sized perimysial vessels. Myofibers in these regions showed moderate fiber size variation,
occasional necrosis, and varied immaturity (Fig. 3A). ATPase
stain at pH 4.3 showed frequent Type IIC fibers. Caveolin-3
cytoplasmic aggregates were common (Fig. 3B). Cytochrome
oxidase stain showed pale areas in occasional fibers (Fig. 3C).
Capillaries were usually enlarged, showed C5-b9 deposition
(Fig. 3D), and increased alkaline phosphatase staining.
Clusters of Regenerating Muscle Fibers
Clusters of regenerating muscle fibers were present in 3
biopsies (Patients 1, 4, 5) often in sections that contained necrotic
fiber clusters in other areas (Fig. 4A). Muscle fibers were small,
round, often Type 2C (Fig. 4B), contained large nuclei, and had
basophilic (Fig. 4A) and alkaline phosphataseY and caveolin3Ypositive cytoplasm (86%), with coarse internal architecture
(Fig. 4C) and MHC-1 expression. Alkaline phosphatase strongly
stained endomysial connective tissue between muscle fibers
(Fig. 4D). Capillaries were enlarged and had varied orientations
and strong alkaline phosphatase staining (Fig. 4D) in qualitatively normal numbers with capillaries adjacent to all muscle
FIGURE 4. Clusters of regenerating muscle fibers. (A) Large
group of small muscle fibers in a similar stage of regeneration
(H&E stain). (B) ATPase pH 4.3 stain shows many immature
intermediate-stained Type 2C muscle fibers. (C) NADH stain
shows coarse internal architecture in small myofibers. (D) Alkaline phosphatase (ALK) is strongly expressed in muscle fibers
and capillaries. Scale bar = (AYD) 50 Km.
fibers. The perimysium and endomysium contained scattered
acid phosphataseYpositive cells.
Anatomic Distribution of Regions of Pathology
The clusters of necrotic muscle fibers (86%) were generally located in ‘‘border zones’’ (Fig. 1A, D), distant from
intermediate-sized perimysial vessels and between their territories. Regions of necrotic fibers often did not have a relationship to the borders of fascicles (Fig. 1D). Purlieu areas were
located between necrotic fiber clusters and intermediate-sized
perimysial vessels. Perimysial vessels were typically located in
the center of purlieu areas (Fig. 1, arrowheads). In areas where
clusters of necrotic muscle fibers were the largest, intervening
purlieu regions were narrow or small (Fig. 1B, right). In 1
patient (No. 6), a cluster of necrotic muscle fibers was located
near a perimysial vessel at the edge of the section, but the
vessel was intact and the purlieu region had normal capillaries.
Ultrastructural Pathology (Patients 2, 4, and 5)
FIGURE 3. Pathology in perivascular purlieu. (A) Fiber size variation and myopathic changes (H&E stain). (B) Intracytoplasmic
aggregates (caveolin-3 immunostain). (C) Cytochrome oxidase
(COX) histochemical stain variably stains myofiber cytoplasm.
(D) Capillaries are mildly enlarged and have C5b-9 deposition.
Scale bar = (AYD) 50 Km.
1130
In regions with necrotic muscle fiber clusters, myofibers
showed large lipid droplets, loss of myofibrils, and accumulation of granular debris (Fig. 5A, B). Capillaries were mostly
lost with only occasional necrotic remnants (Fig. 5A, B, open
arrowheads). Terminal arterioles had obliterated lumens and a
featureless architecture lacking cellular details (Fig. 5A, black
arrow). In purlieu areas, capillaries were abundant and enlarged (Fig. 5C, open arrowhead). Abnormal endothelial cell
features included swelling, large nuclei, a prominent nucleolus, and disjunction (Fig. 5D). Capillary architecture was often disrupted with extravasated red blood cells (Fig. 5D,
arrows). Occasional platelet-rich microthrombi that occluded
capillary lumen were present. No tubuloreticular bodies were
identified in endothelial cells in any of the patients. Muscle
Ó 2014 American Association of Neuropathologists, Inc.
Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.
J Neuropathol Exp Neurol Volume 73, Number 12, December 2014
FIGURE 5. Myofibers and capillaries in necrotic and regenerating
zones. (A) Necrotic zone shows increased space between muscle
fibers, loss of capillaries with occasional outlines of residual architecture (open arrowheads), and myofibers with pale often-vacuolated
cytoplasm and increased lipid droplets. Plastic-embedded, toluidine blueYstained section. (B) Myofiber cytoplasm shows loss of
myofibrillar apparatus, granular debris, and increased lipid droplets (black arrowheads). Necrotic endomysial capillary remnants
with effaced architecture (open arrowhead) are located between
myofibers. Electron photomicrograph. (C) Regenerating zone
shows small myofibers (black arrowheads) with excessive sarcolemmal folds, widened endomysium, and enlarged capillaries
with disrupted architecture (open arrowhead) (plastic-embedded,
toluidine blueYstained section). (Inset) Excessive sarcolemmal
folds in myofiber (electron photomicrograph). (D) Markedly enlarged capillary has disrupted architecture. L, lumen; En, swollen
and extended endothelial cytoplasm; black arrows, extravasated
red blood cells. Electron photomicrograph. Scale bars = (A, C)
10 Km; (B, D) 5 Km.
fiber size was abnormally varied. Many small muscle fibers had
excessive sarcolemmal folding (Fig. 5C, inset black arrowheads), a characteristic feature of muscle fiber atrophy. Other
small fibers had central enlarged nuclei with open chromatin
and prominent nucleoli, features of immaturity. Endomysial
connective tissue appeared mildly increased.
DISCUSSION
The Pattern of Regional Necrosis in RIIM
Suggests a Vascular Etiology
The distinctive pattern of muscle fiber pathology that
defined our series of patients was necrosis or regeneration of
clusters of contiguous muscle fibers. In our patients with an
acquired, probably immune-mediated, regional myopathy, necrotic regions contained muscle fibers in similar early stages of
necrosis with little phagocytic infiltration. The necrotic regions
were usually located in peripheral regions of vascular territories
in the muscle. The arrangement resembles the ‘‘border zone’’
Regional Ischemic Myopathy
patterns in brain pathology that occur with hypoperfusion damage (7) and suggest an ischemic etiology underlying the muscle
fiber damage. Regions of muscle fibers with both necrosis and
hypercontraction occur in limb muscles of experimental animals
subjected to ischemia (8, 9). These models suggest that muscle
fiber changes in the necrotic regions in our patients developed
across periods of days to a few weeks. Regions of muscle fiber
necrosis in patients also included damage to other structures.
Endomysial areas in regions of muscle fiber necrosis were widened and pale, suggesting edema. Muscle fiber and endomysial and perimysial connective tissue expressed the hypoxiainducible marker CA9 (5, 6). Vessel pathology included damage to, and loss of, capillary basal lamina and endothelium and
abnormal alkaline phosphatase staining of some of the few
remaining capillaries. We interpret the loss of endothelium,
C5b-9 deposition, and alkaline phosphatase staining of capillaries in necrotic zones as phenomena that are secondary to
ischemic damage. This is similar to the capillary pathology among
atrophied muscle fibers in childhood dermatomyositis with
vascular pathology (DM-VP) (10). Changes occurring in multiple tissues suggest areas of ischemia rather than a pathologic
process focused only on muscle fibers. The lack of phagocytic
infiltration of most necrotic muscle fibers could be explained
by a lack of reperfusion within the peripheral necrotic zones,
which were most distant from the feeding perimysial artery.
Carbonic anhydrase IX staining of perimysial connective tissue,
thought to be a hypoxia marker in neoplasms, also supports
an ischemic pathogenesis but could also simply be a marker of
damaged perimysium.
Clusters of regenerating, or immature, muscle fibers
typically contained fibers with similar size and state of regeneration that immediately surrounded an intermediate-sized
vessel. These areas resemble regions of reperfusion in experimental animals with limb ischemia in which muscle fibers in
uniform stages of phagocytosis and regeneration develop after
restoration of circulation (8, 9).
Damage to intermediate-sized vessels, and regions of
perimysium surrounding them, supports the idea that a disorder
of larger vessels is associated with the regional myopathy. An
unusual feature of the intermediate-sized vessel pathology is
that veins and their surrounding connective tissues were more
involved than arteries (Fig. 2). Pathology in and around the
walls of intermediate-sized veins could be evidence of primary
damage to these vessels or of abnormalities in larger, more
proximal, vessels with secondary changes in the perimysial veins.
The necrotic zones often simultaneously involved several adjacent perimysial vascular territories, an indication that the impaired
blood flow might have occurred above the level of intermediateto large-sized perimysial feeding vessels. The pathologic alterations of perimysial veins, including prominently enlarged
endothelium, thickened walls, mural and perivascular histiocytic cells, patchy loss of smooth muscle cells in the wall of
veins, and disrupted external elastic lamina and connective tissue, could be a primary pathology and differ from the vascular
abnormalities seen in other immune myopathies. Some of the
cellularity in vein walls could be secondary because postcapillary
venules and veins are the primary sites of leukocyte margination
(11). Vascular pathology seen in other neuromuscular disorders
but not present in RIIM patients included arterial damage,
Ó 2014 American Association of Neuropathologists, Inc.
Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.
1131
J Neuropathol Exp Neurol Volume 73, Number 12, December 2014
Cai et al
necrosis in hereditary myopathies and is not consistent with the
anatomic association with vessels.
The common clinical features in our RIIM patients
comprise a syndrome that can be described as a paraneoplastic
dermatomyopathy. Features include a myopathy with proximal symmetric weakness and a high CK, skin rash, and an
associated malignancy. Our results have parallels to previous
descriptions of paraneoplastic dermatomyositis (14). Patients
with paraneoplastic dermatomyositis syndromes often have
disease onset in the sixth decade or older. Skin necrosis, which
is more common in paraneoplastic dermatomyositis, may be a
counterpart to the, probably ischemic, muscle fiber necrosis in
our RIIM patients. Serum CK is generally higher in paraneoplastic
dermatomyositis syndromes and could reflect an abundance of
necrotic muscle fibers. Serum antibodies to TRIM-33 (TIF-1F)
occur in some adults with paraneoplastic dermatomyositis syndromes (15, 16). Although there are statistical associations
between these antibody-related disorders, and more generally
between adult immune dermatomyopathy syndromes and increased frequency of neoplasms, their associated muscle pathologies have not been described in detail.
Our RIIM patients should be considered in the context
of the common clinical classification of immune myopathies,
which includes polymyositis, dermatomyositis, inclusionbody myositis, and necrotic myopathies compared with our
proposed pathologic groups. Polymyositis and inclusion-body
myositis can be ruled out because there are no mononuclear
cell inflammatory foci or inclusion bodies in RIIM. Necrotic
myopathies are one possible class within which to group our
perimysial vascular fragments, and perivascular B cells in DMVP (10), perivascular mononuclear cell foci in vasculitis, and
neovascularization with clustered smaller vessels.
Capillary pathology in purlieu areas is also consistent
with a vascular disorder in RIIM patients. Purlieu endomysial
capillaries were normal in number but frequently enlarged.
The capillary changes could be related to compensatory vasodilation in the areas outside the region of most severe ischemia.
A similar pattern of endomysial capillary enlargement occurs in
DM-VP in regions near the perifascicular atrophy and is thought
to be caused by intermediate-sized vessel pathology. The ultrastructural changes in capillaries, including endothelial swelling, leukocyte margination, and platelet-rich microthrombi, have
also been described in human ischemic skeletal muscle with
reperfusion (12).
Regional necrosis of muscle fibers has been reported in
several types of conditions. In IMPP, muscle fiber necrosis
and regeneration are often clustered near regions of perimysial
pathology. Pathology within each IMPP region includes muscle
fibers in varied stages of necrosis and regeneration (1, 3). This
differs from RIIM patients in whom fibers within each region
were in a similar stage of pathology. Myopathic grouping with
regions containing muscle fibers in a similar stage of necrosis
or regeneration occurs in some hereditary myopathies, especially
those caused by sarcolemmal disorders such as dystrophinopathies. An early postulate was that myopathic grouping was
caused by vascular pathology (13). However, it now seems more
likely that eccentric contraction of muscle fibers with underlying
fragile sarcolemmal membranes is responsible for the regional
TABLE 3. Dermatomyositis: Histologic Subtypes
DM-VP
Muscle fibers
Pathology
Pathology distribution
Inflammation
Vessel pathology
Capillaries
Perimysial
Intermediate-sized
Perimysial connective tissue
Clinical associations
References
IMPP
RIIM
MDA5 Antibody
Atrophy
COX stain reduced
Caveolin-3 aggregates
Perifascicular
Near avascular perimysium
Location: Perivascular
Type: Lymphocytes
B and T cells
Necrosis
Necrosis
NA
Near perimysium
Regional clusters
Border zones
Location: Veins
Type: Leukocytes
NA
Muscle fiber atrophy region:
Endothelium lost
Purlieu zone: Large
Vascular fragments
Normal
Necrotic zone: Lost
Purlieu zone: Large
NA
Normal
Normal
Fragmented
histiocytic cells
Onset age: Adult 9 child
Skin: Mechanic’s hands
Lungs: Interstitial fibrosis
Weakness
Neoplasm: Rare
Antibodies: tRNA synthetase
Aldolase: May be selectively high
Vein 9 artery
Leukocytes in wall
Normal
Onset age: Child and adult
Skin: Heliotrope rash
Extensor limb surface
Weakness
Calcinosis
Serum CK: Normal or
Mildly high
(1, 10)
Location: Perimysial
Type: Histiocytes
(3)
Onset age: Late adult
Skin: Rash on face, trunk,
and limbs Weakness
Neoplasm: Often
Serum CK:
Often high
Current series
NA
NA
Onset age: Adult
Skin: Ulcers
Palmar papules
Alopecia
Lungs: Interstitial fibrosis
Strength: Normal
Neoplasm: No
Aldolase: May be
selectively high
(19)
CK, creatine kinase; COX, cytochrome oxidase; DM-VP, dermatomyositis with vasculopathy; IMPP, immune myopathy with perimysial pathology; NA, not applicable; RIIM,
regional ischemic immune myopathy.
1132
Ó 2014 American Association of Neuropathologists, Inc.
Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.
J Neuropathol Exp Neurol Volume 73, Number 12, December 2014
patients, but such a classification would be misleading. Classifying disorders as ‘‘necrotic’’ generally suggests that the underlying pathogenic factors, like toxins, metabolic or immune
disorders, or infections directly damage the necrotic cell type
(17). Although our patients were chosen because their pathology included necrotic muscle fibers, our data suggest that the
underlying lesion is vascular. By analogy, a classification scheme
that referred to RIIM as a ‘‘necrotic myopathy’’ would inappropriately refer to ischemic cerebrovascular disorders as ‘‘necrotic
encephalopathies’’ or myocardial infarcts as ‘‘necrotic cardiomyopathies.’’ Another possibility is to group RIIM as a form of
dermatomyositis, a term that denotes the co-occurrence of inflammatory myopathy and skin pathology; however, no focal
inflammation was observed in our RIIM patients. Furthermore,
the term ‘‘dermatomyositis’’ lacks pathologic specificity. There
now appears to be at least 4 clinically, immunologically, and
pathologically different immune clinical dermatomyopathy syndromes (Table 3). None of the others has the regional muscle
fiber necrosis and regeneration with venous pathology seen in
RIIM. Features in DM-VP but not RIIM include vessel pathology with perivascular mononuclear cell foci often containing B cells and fragmented intermediate-sized perimysial
vessels (10) and muscle fiber morphologic alterations such as
atrophy, vacuoles, and mitochondrial pathology (18), without
necrosis or C5b-9 deposition in fiber cytoplasm. The distribution of muscle fiber atrophy in DM-VP is linear, located most
prominently near the perimysium, whereas the clustered necrosis in RIIM occurs in large regions in border zones between
artery-vein pairs, showing little relation to the distribution of
the perimysium. Dermatomyositis with vascular pathology is
commonly a childhood disorder in contrast to RIIM. Dermatomyositis syndromes with IMPP pathology have damage to
and diffuse presence of histiocytic cells in perimysial connective tissue (3), that is, changes that were not present in RIIM
patients. Necrosis is common in IMPP syndromes, but regions
with muscle fiber pathology contain fibers in different stages
of necrosis and regeneration rather than the regionally uniform
stages of necrosis seen in RIIM patients. IMPP, but not
RIIM, syndromes are commonly associated with interstitial
lung disease. A fourth dermatomyositis syndrome associated
with MDA5 antibodies has mild muscle involvement and a
distinctive palmar rash on flexor surfaces of the hands (19),
which was not seen in RIIM patients. Association with neoplasms is uncommon in DM-VP, IMPP, and MDA5
antibodyYassociated syndromes.
We suggest that RIIM pathology with border zone distribution of muscle fiber and capillary damage and CA9 staining
is most consistent with a pathologic classification as a vasculopathy with secondary regional ischemia and muscle fiber
damage. Vascular pathology was present in veins but not arteries in all patients. Understanding the relationship between the
vein pathology and ischemia remains for further investigations.
Additional studies to clarify the pathophysiology of RIIM include evaluation of the skin to look for vascular pathology, serum for associated autoantibodies, and muscle vasculature to
define the vessel pathology causing the border zone ischemia
in muscle. The clinical syndrome in RIIM can be described as
a paraneoplastic dermatomyopathy syndrome in older patients.
Regional Ischemic Myopathy
Although the myopathology lacks mononuclear cell foci or
humoral immune features, the rapid onset of a diffuse symmetric
myopathy and response to corticosteroid treatment in some
patients suggest that the syndrome is immune mediated. A series of paraneoplastic dermatomyopathy patients with associated serum antibody binding to nuclear matrix protein NXP-2 or
transcription intermediary factor 1F has been described (20),
but no detailed clinical or myopathologic information was
presented. Studies of additional RIIM patients are required to
define further the range of clinical and pathologic features
and any associated serum autoantibodies.
REFERENCES
1. Pestronk A. Acquired immune and inflammatory myopathies: Pathologic
classification. Curr Opin Rheumatol 2011;23:595Y604
2. Levin MI, Mozaffar T, Al-Lozi MT, et al. Paraneoplastic necrotizing
myopathy: Clinical and pathological features. Neurology 1998;50:764Y67
3. Mozaffar T, Pestronk A. Myopathy with anti-Jo-1 antibodies: Pathology
in perimysium and neighbouring muscle fibres. J Neurol Neurosurg
Psychiatry 2000;68:472Y78
4. Miller T, Al-Lozi MT, Lopate G, et al. Myopathy with antibodies to the
signal recognition particle: Clinical and pathological features. J Neurol
Neurosurg Psychiatry 2002;73:420Y28
5. Lal A, Peters H, St Croix B, et al. Transcriptional response to hypoxia in
human tumors. J Natl Cancer Inst 2001;93:1337Y43
6. Wykoff CC, Beasley NJ, Watson PH, et al. Hypoxia-inducible expression
of tumor-associated carbonic anhydrases. Cancer Res 2000;60:7075Y83
7. Belden JR, Caplan LR, Pessin MS, et al. Mechanisms and clinical features of posterior border-zone infarcts. Neurology 1999;53:1312Y18
8. Sonmez TT, Al-Sawaf O, Brandacher G, et al. A novel laser-Doppler
flowmetry assisted murine model of acute hindlimb ischemia-reperfusion
for free flap research. PloS One 2013;8:e66498
9. Carmo-Araujo EM, Dal-Pai-Silva M, Dal-Pai V, et al. Ischaemia and
reperfusion effects on skeletal muscle tissue: Morphological and histochemical studies. Int J Exp Pathol 2007;88:147Y54
10. Pestronk A, Schmidt RE, Choksi R. Vascular pathology in dermatomyositis
and anatomic relations to myopathology. Muscle Nerve 2010;42:53Y61
11. Muller WA. Mechanisms of leukocyte transendothelial migration. Ann
Rev Pathol 2011;6:323Y44
12. Blaisdell FW. The pathophysiology of skeletal muscle ischemia and the
reperfusion syndrome: A review. Cardiovascular Surg 2002;10:620Y30
13. Mendell JR, Engel WK, Derrer EC. Duchenne muscular dystrophy:
Functional ischemia reproduces its characteristic lesions. Science 1971;
172:1143Y45
14. Fardet L, Dupuy A, Gain M, et al. Factors associated With underlying
malignancy in a retrospective cohort of 121 patients with dermatomyositis. Medicine (Baltimore) 2009;88:91Y97
15. Trallero-Araguás E, Rodrigo-Pendás JA, Selva-O’Callaghan A, et al.
Usefulness of anti-p155 autoantibody for diagnosing cancer-associated
dermatomyositis: A systematic review and meta-analysis. Arthritis Rheum
2012;64:523Y32
16. Fujimoto M, Hamaguchi Y, Kaji K, et al. Myositis-specific anti-155/140
autoantibodies target transcriptional intermediary factor 1 family proteins. Arthritis Rheum 2012;64:513Y22
17. Aronica E, Stenzel W. Reactins of muscle to toxins and drugs. In: Goebel
H-H, Sewry CA, Weller RO. Muscle Disease Pathology and Genetics,
International Society of Neuropathology; Ebrary Medical Subscription
Collection. 2nd ed. Chichester, West Sussex, UK: Wiley Blackwell, 2013:
328Y38
18. Varadhachary AS, Weihl CC, Pestronk A. Mitochondrial pathology in immune and inflammatory myopathies. Curr Opin Rheumatol 2010;22:651Y57
19. Fiorentino D, Chung L, Zwerner J, et al. The mucocutaneous and systemic phenotype of dermatomyositis patients with antibodies to MDA5
(CADM-140): A retrospective study. J Am Acad Dermatol 2011;65:25Y34
20. Fiorentino DF, Chung LS, Christopher-Stine L, et al. Most patients with
cancer-associated dermatomyositis have antibodies to nuclear matrix
protein NXP-2 or transcription intermediary factor 1F. Arthritis Rheum
2013;65:2954Y62
Ó 2014 American Association of Neuropathologists, Inc.
Copyright © 2014 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.
1133