Canine Pericardial Mesothelioma - PDF by pengtt


									             Veterinary Pathology Online 

                                          Canine Pericardial Mesothelioma
                                 S. P. McDonough, N. J. MacLachlan and A. H. Tobias
                                               Vet Pathol 1992 29: 256
                                         DOI: 10.1177/030098589202900312

                                    The online version of this article can be found at:

                                                                Published by:


                                                                 On behalf of:
American College of Veterinary Pathologists, European College of Veterinary Pathologists, & the Japanese College of Veterinary

                      Additional services and information for Veterinary Pathology Online can be found at:

                                          Email Alerts:




                                               Downloaded from by guest on December 6, 2010
Vet Pathol 29256-260 (1992)

                                         Canine Pericardial Mesothelioma
                          S. P. MCDONOUGH, J. MACLACHLAN, A. H. TOBIAS
                                        N.             AND

      Key words: Dogs; mesothelioma; pericardium.

   Mesothelioma of the canine pericardium has been de-                      and a large irregular multilobular mass that was firmly ad-
scribed infrequently, usually as one manifestation of wide-                 herent to the diaphragm, epicardium, and the right pleural
                                               . ~ ~ report
spread dissemination of the n e o p l a ~ m This ~ ~ ~ de-                  surface. Neoplastic nodules were also present in the lung and
scribes pericardial lesions in three dogs that were presented               tracheobronchial lymph nodes.
to the Veterinary Medical Teaching Hospital (University of                       1
                                                                               A 1 tissues collected after pericardiectomy or necropsy were
California, Davis, CA) because of recurrent pericardial ef-                 immersed in 10% neutral buffered formalin and embedded
fusion. All three dogs had been treated with repeated peri-                 in paraffin. Sections were cut and stained with hematoxylin
cardiocentesis and eventually by subtotal pericardiectomy.                  and eosin. The subserosal connective tissue layer of all peri-
Histologic examination of the resected pencardial sacs from                 cardial sacs contained multiple nodules of large cells arranged
all three dogs revealed nests of mesothelial cells. All three               in solid nests or irregular cords and acini that were bounded
dogs died within 1 month of surgery, and necropsy of one                    by a scant to moderate amount of fibrovascular stroma (Fig.
dog confirmed fulminant metastatic mesothelioma subse-                       1). The pericardial sac from the Airedale Terrier (case No.
quent to pericardiectomy. The other two dogs were unavail-                   1) also had multiple nodules of glandular elements that pro-
able for postmortem examination.                                            jected in small dome-like masses above the surface of the
   The three dogs included an 11-year-old male Airedale Ter-                serosal layer. Cells were cuboidal to polygonal with a large
rier (case No. l), a 9-year-old castrated male Golden Re-                   oval nucleus that contained coarsely stippled to vesicular
triever (caseNo. 2), and an 11-year-old Airedale Terrier cross              chromatin and one to two large prominent nucleoli. The cells
(caseNo. 3). All three dogs had signs ofright-sided congestive              had abundant amounts of granular eosinophilic cytoplasm
heart failure characterized by ascites and distended jugular                and distinct cell borders, and one to three mitotic figures
veins. Radiographic examination revealed globular cardiac                   were present per high-power field (500 pm in diameter). Dif-
silhouettes; examination by electrocardiographyshowed small                 ferential diagnoses included metastatic squamous cell car-
complexes and electrical alternans, and echocardiographic                   cinoma, adenocarcinoma, and mesothelioma. The histologic
examination confirmed pericardial effusion. No masses as-                   appearance of cells within neoplastic nodules collected at
sociated with the right atrium, aorta, or other structures with-            necropsy of the Golden Retriever (case No. 2) was identical
in the pericardial space were visualized during the ultrasound              to that of cells within the pericardium, but rafts of cells were
examination. The Airedale Terrier (case No. 1) was treated                  frequently present within endothelium-lined vessels.
with repeated pericardiocentesis over a 4-month period, and                    Ultrastructural analysis and immunohistochemical stain-
the pericardial effision of the Golden Retriever (case No. 2)               ing of the tumor cells were undertaken to confirm their his-
was drained twice over the course of 1 year. The Airedale                   togenic origin. The presence of both cytokeratins and vi-
Terrier cross (case No. 3) had a single bout of pericardial                 mentin in tumor cells within the pericardial sacs of affected
effusion in 1985 that recurred 5 years later and was then                   dogs was established by avidin biotin peroxidase complex
drained twice. All these dogs subsequently underwent medial                 method staining using commercially available monoclonal
sternotomy for exploration of the thoracic cavity and subtotal              antibodies that were specific for high- and low-molecular-
pericardiectomy. Neoplasms were not detected in any of the                  weight cytokeratins (AE 1/AE3, Boehringer Mannheim, In-
dogs during exploratory thoracotomy. Several weeks after                    dianapolis, IN), primarily low-molecular-weight cytokera-
surgery, all three dogs developed marked pleural effusion.                  tins (PKK1, Labsystems, Raleigh, NC and CAM5.2, Becton
The owners of the Airedale Terrier (case No. 1) and Airedale                Dickinson, Mountain View, CA), high-molecular-weight cy-
Terrier cross (case No. 3) declined further treatment, and                  tokeratins (A575, DAKO Corp., Carpinteria, CA), and vi-
subsequently both dogs died but were unavailable for nec-                   mentin (ICN Biochem, Lisle, IL). Factor VIII-related antigen
ropsy. The Golden Retriever (case No. 2) was treated with                   staining (Calbiochem Corp., San Diego, CA) was done in an
repeated thoracocentesis and two doses of intracavitary Cis-                attempt to demonstrate vascular invasion by neoplastic cells.
platinum@for suspected pleural mesothelioma. The dog re-                    All staining was done using the standard avidin biotin per-
sponded poorly to therapy and was euthanatized. At nec-                     oxidase complex method of Hsu et al.’ and the Vectastain
ropsy, this dog had a moderate serosanguinouspleural effision               kit (Vector Laboratories, Burlingame, CA). All tests were

  Fig. 1. Pericardium; dog No. 3. Nests of apparently neoplastic cells are surrounded by fibrovascular stroma. HE. Bar
=  50 pm.
  Fig. 2. Pericardium; dog No. 3. Pancytokeratin (AEUAE3) staining is intense and distributed diffusely throughout the
cytoplasm of apparently neoplastic cells in Fig. 1. Avidin biotin peroxidase complex method, Mayer’s hematoxylin coun-
terstain. Bar = 100 pm. Inset: Same section stained for vimentin. Note positively stained cells at the arrow. Bar = 10 pm.

                                             Downloaded from by guest on December 6, 2010
Vet Pathol 29:3, 1992                      Brief Communications and Case Reports                                        251

  Fig. 3. Pleural mass; dog No. 2. Pancytokeratin (AEUAE3) staining is intense in neoplastic cells. Avidin biotin peroxidase
complex method, Mayer’s hematoxylin counterstain. Bar = 50 pm.
  Fig. 4. Pleural mass; dog No. 2. Vimentin staining of the same tissue as in Fig. 3. Note clusters of intensely staining
cells at the arrows. Avidin biotin peroxidase complex method, Mayer’s hematoxylin counterstain. Bar = 50 pm.
   Fig. 5. Electron micrograph. Pericardium; dog No. I. Tumor cell in subserosal nest. Note microvilli around the entire
circumference of the cell, numerous well-developed desmosomes (arrow), and prominent intercellular spaces. Bar = 2 pm.
  Fig. 6. Electron micrograph. Pleural mass; dog No. 2. Note well-developed microvilli with primary and secondary
branching. Bar = 1 pm.

                                           Downloaded from by guest on December 6, 2010
258                                             Brief Communications and Case Reports                                                   Vet Pathol293, 1992

                              Table 1. Summary of immunohistochemical staining of neoplasms.
         Tumor Type          Case No.                  Breed                                 Sex*               Age               AE 1/AE3
                                                                                                                         Intensityt       Pattern$
      Pericardial                1         Airedale Terrier                     M               11                        +++
        mesotheliomas            2         Golden Retriever                     M/c              9
                                             biopsy material                                                              +++
                                             necropsy material                                               ++          Diffise
                                   3       Airedale Terrier cross               M               11          +++          Diffuse
     Pulmonary                    4        Cockapoo                             M/C             16           ++          Peripheral
       adenocarcinomas             5       Doberman Pinscher                    F/s           > 16          +++          Peripheral
                                  6        Shetland Sheepdog                    M               12                        +++
     Epithelial type              7        Yorkshire Terrier                    F               13                        +++
       mesotheliomas              8        German Shepherd Dog                  M/C             14          +++          Diffuse
                                   9       German Shepherd Dog                  M/c             15           ++          Diffise
                                 10        Labrador Retriever                   M                7          +++          Diffuse
                                 11        Terrier cross                        F/s             10                          +
     Chemodectomas 11            12        Labrador Retriever                   F/s          > 16             0
                                 13        Australian Shepherd Dog              F/s             11            0
                                 14        Golden Retriever                     F/s             14            0
  * M = intact male; M/c = castrated male; F/s= spayed female.
  t ++ + = most cells stained with many staining intensely; + + = many cells stained moderately and some stained intensely; + = many
cells stained weakly and some stained moderately or intensely; 0 = no cells stained.
     Diffuse = the reaction product was evenly distributed throughout the cytoplasm; peripheral = the reaction product was confined to the
periphery of the cytoplasm.
  9 ND = not done.
  11 Chemodectomas were weakly positive for Chromogranin A but negative for neuron-specific enolase.

performed on formalin-fixed paraffin-embedded tissue with                      react moderately to intensely with both high- and low-mo-
3,3’-diaminobenzidine (Sigma, St. Louis, MO) as the chro-                      lecular-weight cytokeratins. The chemodectomas were neg-
mogen. All cytokeratin-stained slides were predigested with                    ative for both cytokeratins and vimentin, weakly positive for
0.1% trypsin (Sigma, St. Louis, MO) for 20 minutes before                      Chromogranin A, and negative for neuron-specific enolase
the primary antibody was applied. The staining pattern of                      (data not shown). The staining pattern of the adenocarci-
the cells within the affected pericardial sacs was compared                    nomas was distinctive, with a markedly enhanced peripheral
with histologically confirmed cases of canine pulmonary ad-                    (membrane) region imparting a ring-like appearance to the
enocarcinoma (n = 3), disseminated epithelial-type meso-                       cells. This staining pattern was not seen to an equivalent
thelioma (n = 5), and chemodectoma (n = 3), as summarized                      extent in any of the other types of neoplasms.
in Table 1. The pericardial sac tumors and the chemodec-                          Selected thin sections of affected pericardial sacs and met-
tomas were further evaluated with Chromogranin A (Incstar,                     astatic tumors collected at necropsy from each case were
Stillwater, MN) and neuron-specific enolase (DAKO Corp.,                       stained with 2%uranyl acetate in 95% alcohol for 16 minutes
Santa Barbara, CA). Negative technique controls were per-                      and in Reynolds’ lead citrate for 9 minutes and examined
formed by substituting the primary antibody with phosphate-                    with a Zeiss transmission electron microscope. Ultrastruc-
buffered saline and also by using normal rabbit serum. Pos-                    tural details of the neoplastic cells were distinct (Figs. 5, 6).
itive and negative tissue controls consisted of a block of 2 1                 Numerous long microvilli with primary and secondary
normal canine tissues.                                                         branching surrounded the entire circumference of the cells.
   Tumor cells within the affected pericardial sacs coex-                      Adjacent neoplastic cells were joined by frequent well-de-
pressed vimentin as well as both high- and low-molecular-                      veloped desmosomes, and intercellular spaces were promi-
weight cytokeratins that typically were distributed diffusely                  nent. Cells had abundant cytoplasm that contained numerous
throughout the cytoplasm (Fig. 2). Neoplastic cells in tissues                 bundles of tonofilaments arranged circumferentially around
taken at necropsy from the Golden Retriever (case No. 2)                       the nucleus, rough endoplasmic reticulum, scattered glycogen
had the same staining reactivity (Figs. 3, 4). The epithelial-                 granules, and moderate numbers of mitochondria. Sections
type mesotheliomas usually had a similar staining pattern,                     removed at surgery from the pericardium of the Golden Re-
except vimentin staining was restricted to scattered cells.                    triever (case No. 2) also contained cells with numerous mi-
Whereas the epithelial-type mesotheliomas inconsistently re-                   crovilli, but autolysis prevented their definitive character-
acted to CAM5.2, both pericardial sac tumors on which this                     ization. The ultrastructural characteristics of neoplastic cells
staining was done reacted moderately to intensely. None of                     collected from metastatic tumors at necropsy from this dog
the pulmonary adenocarcinomas expressed vimentin but did                       were compatible with those of mesothelioma.

                                                Downloaded from by guest on December 6, 2010
Vet Pathol 29:3, 1992                        Brief Communications and Case Reports                                                         259

                                                         Table 1. Extended.

                        PKKl                            CAM5.2                                                 HMW             Intensity
     Intensity                 Pattern    Intensity                  Pattern                       Intensity         Pattern
      +++                  Peripheral       ++                  Peripheral                           +++             Diffuse     +
      ++                   Peripheral                           NDO                                  +++             Diffise     +
      +++                  Diffise                              ND                                   +++             Diffuse     ++
      +++                  Diffuse          ++                  Diffuse                              +++             Diffuse      +
      ++                   Peripheral        0                                                                       ND           0
      +++                  Peripheral       ++                   Peripheral                                          ND           0
      +++                  Peripheral       ++                   Peripheral                                          ND           0
      +++                  Diffuse          +                    Diffuse                                             ND           0
      +++                  Diffuse           0                                                                       ND           +
      +++                  Diffuse           0                                                                       ND           +
      +++                  Diffuse           +                   Diffuse                                             ND           +
      +++                  Diffuse           0                                                                       ND           +
         0                                                       ND                                                  ND           0
         0                                                       ND                                                  ND           0
         0                                                       ND                                                  ND           0

   Pericardial effusion is an infrequent cause of cardiovas-                subserosal cells, but the possibility of derivation from the
cular disease in the dog.' Reported etiologies include neo-                 surface mesothelium with dedifferentiation cannot be ex-
plasia, benign idiopathic pericardial effusion, primary car-                cluded. The ultrastructural features of the neoplastic cells
diac disease, trauma, infection, and uremia. The three dogs                 also were characteristic of mesothelial rigi in.^.^.^,^.'^
described in this report had a distinctive syndrome of peri-
cardial effusion associated with mesothelial cell proliferation,
which in one case progressed to disseminated mesothelioma.
Because of similarities in the postsurgical outcomes, we be-                  We thank D. Naydan for her enthusiastic assistance with
lieve the mesothelial proliferations in the two remaining dogs              the immunohistochemistry and Drs. C. McCullough, D.
also indicated mesothelioma, although their malignant po-                   Danilenko, and B. Rideout for their original descriptions of
tential could not be confirmed because they were not nec-                   the biopsy material. We also thank Dr. R. J. Higgins for his
ropsied. No direct evidence of lymphatic invasion was found                 helpful discussions.
by examination with light microscopy or electron microscopy
or by the use of Factor VIII-related antigen staining to iden-
tify lymphatic vessels. The pericardial effision was most like-                                                References
ly secondary to increased hydrostatic pressure from lym-                      1 Berg RJ, Wingfield W Pericardial effision in the dog: a
phatic obstruction. The rapid postsurgical deterioration in                     review of 42 cases. J Am Hosp Assoc 20:721-730, 1984
all three dogs likely resulted fiom transthoracic spread of                   2 Bolen JW, Hammar SP, McNutt MA: Reactive and neo-
neoplastic mesothelial cells at the time of subtotal pericar-                   plastic serosal tissue; a light-microscopic,ultrastructural,
diectomy.                                                                       and immunocytochemical study. Am J Surg Pathol 10:
   Immunohistochemical staining clearly identified these tu-                    34-47, 1986
mors as mesotheliomas and not carcinomas. Normal serosal                      3 Cibas ES, Corson JM, Pinkus GS: The distinction of
tissues consist of a surface mesothelium derived from mul-                      adenocarcinoma from malignant mesothelioma in cell
tipotential subserosal cells in the underlying layer of vascu-                  blocks of effusions: the role of mucin histochemistry and
larized connective           Resting multipotential subserosal                  immunohistochemical assessment of carcinoembryonic
cells express only vimentin but coexpress vimentin and low-                     antigen, keratin proteins, epithelial membrane antigen,
molecular-weight cytokeratins as they proliferate. As these                     and milk fat globule-derived antigen. Hum Pathol 18:
cells differentiateinto surface mesothelium, they may express                   67-74, 1987
high-molecular-weight cytokeratins but cease to express vi-                   4 Corson JM: Pathology of malignant mesothelioma. In:
m e n t h 6 Although expression of cytokeratins and vimentin                    Asbestos-related Malignancy, ed. Antman and Aisner,
by mesotheliomas is variable, often reflecting the histologic                   pp. 179-199. Grune and Stratton, Inc., Orlando, FL,
type of tumor, immunohistochemical examination is increas-                      1986
ingly being used to diagnose mesotheliomas.24JoThe peri-                      5 Harbison ML, Godleski JJ: Malignant mesothelioma in
cardial mesotheliomas in this study consistently expressed                      urban dogs. Vet Pathol 20531-540, 1983
vimentin and both low- and high-molecular-weight cytoker-                     6 Holt JP: The normal pericardium. Am J Cardiol 26:
atins. These cells may have been derived from multipotential                    455-465, 1970

                                             Downloaded from by guest on December 6, 2010
260                                           Brief Communication.s and Case Reports                                        Vet Pathol 293, 1992

 7 Hsu SM, Raine L, Fanger H: The use of antiavidin an-                       10 Wick MR, Loy M, Mills SE, Legier JF, Manivel JC:
   tibody and avidin-biotin-peroxidase complex in im-                            Malignant epithelioid mesothelioma versus peripheral
   munoperoxidase technics. Am J Clin Path01758 16-82 1,                         pulmonary adenocarcinoma: a histochemical, ultrastruc-
    1981                                                                         tural, and immunohistologic study of 103 cases. Hum
 8 Ikede BO, Zubaidy A, Gill C W Pericardial mesotheli-                          Pathol 21:759-766, 1990
   oma with cardiac tamponade in a dog. Vet Pathol 17:
   496-499, 1980                                                             Request reprints from Dr. S. P. McDonough, Department of
 9 Trigo FJ, Morrison WB, Breeze R G An ultrastructural                      Veterinary Pathology, School of Veterinary Medicine, Uni-
   study of canine mesothelioma. J Comp Pathol 91:53 1-                      versity of California, Davis, CA 95616 (USA).
   537, 1981

Vet Pathol29260-262 (1992)

               Desmin and Vimentin Immunocharacterization of Feline Muscle Tumors
                            J. MART~N LAS MULAS, H. Vos, AND F. N. VANMIL
                                    DE         J.
      Key words: Cats; desmin; feline tumors; immunohistochemistry; intermediate filaments.

   The immunohistochemical identification of different in-                   feline tissues (skeletal muscle, intestinal smooth muscle, ten-
termediate filaments is a useful aid in the histogenetic clas-               don, and spinal cord) and in 24 histologically nonmuscle
sification of undifferentiated tumors because the intermedi-                 feline soft tissue tumors: neurofibrosarcomas, fibrosarcomas,
ate filament profile of mammalian cells is rather tissue                     lymphosarcomas, and undifferentiated sarcomas (Table 1).
specificloand is generally preserved during neoplastic trans-                Tissue samples had been fixed in 10% neutral buffered for-
f ~ r m a t i o n Some of the commercially available antisera to             malin for 24 to 36 hours and embedded in paraffin. Com-
intermediate filament proteins can be used on formaldehyde-                  mercially available polyclonal antisera directed against
fixed, paraffin-embedded material,I3 allowing retrospective                  chicken gizzard desmin, calf lens vimentin, human spinal
studies to be performed on routinely processed tissue spec-                  cord glial fibrillary acidic protein, and a monoclonal anti-
imens. Desmin is an intermediate filament protein present                    serum raised against human brain neurofilament proteins
in cardiac and skeletal muscle and also in smooth muscle of                  were obtained from Euro-Diagnostics B.V. (Apeldoorn, The
viscera and v e s ~ e l s . ~ J ~ immunohistochemical demon-                 Netherlands). The peroxidase anti-peroxidase and indirect
stration of desmin in tumor cells is a reliable method to                    immunoperoxidase techniques were applied to polyclonal
elucidate the myogenic character of soft tissue tumors in                    and monoclonal antisera, respectively, as described else-
human beings.12 Rhabdomyosarcomas have histologically                        where.13 All primary antibodies were incubated overnight at
variable histomorphologic appearances8and consequently can                   4 C and diluted 1 :80 (polyclonals) and 1 : 10 (monoclonal)
be difficult to diagnose, particularly in the absence of distinct            in phosphate-buffered saline, pH 7.6.13 Normal skeletal and
histomorphologic features indicating rhabdomyoblastic dif-                   smooth muscle tissue sections were used as positive controls.
ferentiation. In human beings, desmin approaches absolute                    Negative controls included the substitution of the specific
sensitivity and specificity as a marker for muscle cell differ-              primary antibodies by both phosphate-buffered saline and
entiation in rhabdomyosarcomas of the embryonal and al-                      normal rabbit or mouse sera diluted 1 :80 or 1 : 10, respec-
veolar types and is also a sensitive marker for pleomorphic                  tively. Counterstaining of immunoperoxidase-treated sec-
rhabdomyosarcoma.l* Desmin labeling of rhabdomyosar-                         tions was performed with Hams’ hematoxylin; tissue sec-
comas in several other species has been reported previously:                 tions were incubated for 1 minute.
in dogs,1.6    rats,ll and mice.5 In cats, the myogenic character               Histologically, all seven rhabdomyosarcomas showed in-
of a presumed rhabdomyosarcoma could not be demonstrat-                      filtrative growth, although two appeared to be partially en-
ed immunohistochemically by labeling for desmin.’                            capsulated. All tumors contained sheets or nests of various
   Seven feline soft tissue tumors histologically diagnosed as               sizes of highly pleomorphic round polyhedral to spindle-
rhabdomyosarcomas were typed immunohistochemically for                       shaped cells, with eosinophilic cytoplasm and a broad range
desmin, vimentin, glial fibrillary acidic protein, and neuro-                in the nucleus: cytoplasm ratio (Fig. 1). Densely packed in-
filament proteins. All cats were European short hairs. The                   terlacing bundles of fusiform cells with eosinophilic cyto-
tumors were at different locations, i.e., at the larynx, mam-                plasm and blunt-ended uniform nuclei sometimes predom-
mary glands M5 (inguinalkM4 (lower abdominal), the ear                       inated. A variable number of round to angular pleomorphic
region, and the thoracioabdominal wall and diaphragm, with                   cells with either one or two bizarre nuclei or several smaller
metastases to the lung and kidney. The same intermediate                     ones was present in all cases. Striations were not seen after
filament profile was also analyzed in histologically normal                  phosphotungstic hematoxylin staining, mitoses were not par-

                                               Downloaded from by guest on December 6, 2010

To top