h-Caldesmon in leiomyosarcoma and tumors with smooth muscle cell-like differentiation: its specific expression in the smooth muscle cell tumor

h-Caldesmon (h-CD) is a protein combined with actin and tropomyosin that regulates cellular contraction. h-CD has been thought to be expressed exclusively in vascular and visceral smooth muscle cells (SMC). We examined h-CD expression immunohistochemically in tumors with SMC and SMC-like differentiation to clarify whether h-CD is specifically expressed in SMC tumors. The tumors examined in this study were six leiomyomas (LM), two angioleiomyomas (ALM), six leiomyosarcomas (LMS), eight rhabdomyosarcomas (RMS), eight malignant fibrous histiocytomas (MFH), four desmoids, three glomus tumors (GT), and two inflammatory myofibroblastic pseudotumors (IMP) of urinary bladder. We found that LM, ALM, LMS, and GT showed intense and extensive immunoreactivity for h-CD, whereas other tumors were completely negative for h-CD. In addition, h-CD was not present in the vascular pericytes and myofibroblasts, in contrast to actin. Although myoepithelial cells were immunopositive for h-CD, neoplastic myoepithelial cells of myoepithelial tumors and mixed tumors of the salivary gland and skin were all negative. These findings indicate that h-CD is a specific marker of both SMC and its neoplasms and that immunohistochemical detection of h-CD may facilitate the differential diagnosis between LMS and other tumors with SMC-like differentiation, including myofibroblastic tumors.     

Calponin and h-caldesmon in soft tissue tumors: consistent h-caldesmon immunoreactivity in gastrointestinal stromal tumors indicates traits of smooth muscle differentiation.

Currently, the immunohistochemical evaluation of smooth muscle differentiation is usually based on desmin, which also reacts with skeletal muscle and is not present in all smooth muscle tumors, and alpha-smooth muscle actin, which reacts with myoepithelial cells. Neither marker typically reacts with gastrointestinal stromal tumors (GISTs), previously classified as smooth muscle tumors or presently often classified as smooth muscle/stromal tumors. Two cytoskeleton-associated actin-binding proteins, calponin (CALP) and h-caldesmon (HCD), are putative smooth muscle markers that also react with myoepithelia. These markers are of particular interest in the immunohistochemical analysis of tumors; neither of them has been extensively documented in soft tissue tumors. In this study, we evaluated selected normal and reactive tissues and more than 250 mesenchymal tumors for CALP and HCD. Both markers were expressed in parenchymal and vascular smooth muscle cells in various organs and in myoepithelial cells. CALP also reacted with myofibroblasts of desmoplastic stroma. All of our 25 benign smooth muscle tumors from various locations were positive for CALP and HCD, as were most of the retroperitoneal and uterine leiomyosarcomas. HCD was more specific, because CALP also reacted with myofibroblastic lesions. The common reactivity of malignant fibrous histiocytomas with CALP and HCD suggests a combination of myofibroblastic and smooth muscle differentiation in these tumors. The GISTs (c-kit positive, usually actin negative) showed nearly consistent HCD reactivity, suggesting traits of smooth muscle differentiation. GISTs were usually CALP negative and showed a CALP expression pattern similar to that of alpha-smooth muscle actin. Although nonmuscle, nonmyofibroblastic tumors were negative for CALP and HCD, synovial sarcomas showed streaks of CALP-positive cells of unknown significance. CALP and HCD should be explored as markers to identify myofibroblastic and smooth muscle cell differentiation in mesenchymal tumors.     

Differential expression of high molecular weight caldesmon in colorectal pericryptal fibroblasts and tumour stroma

AIM: To investigate an extent of smooth muscle differentiation of pericryptal fibroblasts. METHODS: Expression of high molecular weight caldesmon (h-CD) and alpha smooth muscle actin was investigated in 123 invasive colorectal adenocarcinomas and their surrounding non-neoplastic tissues. RESULTS: The monoclonal antibody to h-CD, which showed predominantly positive immunostaining in well differentiated smooth muscle cells, recognised pericryptal fibroblasts, smooth muscle cells, and pericytes, but was not reactive to myofibroblasts. Antibody to alpha smooth muscle actin recognised not only pericryptal fibroblasts, smooth muscle cells, and pericytes but also myofibroblasts. CONCLUSIONS: Pericryptal fibroblasts show greater smooth muscle differentiation than myofibroblasts and there is a possibility that they are well differentiated smooth muscle cells; h-CD is an excellent marker to discriminate pericryptal fibroblasts from myofibroblasts.     

Atypical lipomatous tumors with smooth muscle differentiation: report of two cases

Two cases of soft-tissue atypical lipomatous tumors with areas of smooth muscle cell differentiation are presented. The patients were a 72-year-old man (Case 1) and a 62-year-old woman (Case 2); their neoplasms were located in the left inguinal region and on the chest, respectively. The adipocytic component in Case 1 contained cells with nuclear atypism and lipoblasts. Case 2 showed only slight nuclear atypism and lacked any lipoblasts. Smooth muscle fascicles were composed of well-differentiated cells with a few mitoses having none or only a slight atypism. They expressed muscle-specific actin and desmin in both cases. No recurrence was recorded during the follow-up time.     

Epstein-Barr virus–associated smooth muscle tumor—report of 3 tumors including 1 intracerebral case with a prominent intratumoral B-lymphocytic component and plasma cells

Three Epstein-Barr virus–associated smooth muscle tumors (intracranial, endobronchial, and paraspinal) in 2 patients are presented. The patient with the intracranial tumor had no concurrent or previous manifestation of this neoplasm and was immunosuppressed because of a renal transplant. The other patient had AIDS and radiologic evidence of involvement of multiple other sites. Although mitotic activity was present, none of the tumors displayed any high-grade malignant histologic features. All tumors were strongly positive for smooth muscle actin and variably expressed h-caldesmon. Desmin was negative in all 3 tumors. The intracranial tumor was associated with prominent intracerebral edema and displayed, in addition to a prominent intratumoral T-cell component, a hitherto unreported component of prominent intratumoral B cells and plasma cells.     

Rat hepatic lipocytes express smooth muscle actin upon activation in vivo and in culture.

Myofibroblasts are mesenchymal cells that are prominent in liver injury. The origin of myofibroblasts in liver is debated, although morphologic evidence to date has suggested that these cells are derived from lipocytes (fat-storing cells, Ito cells). In the present study, we have utilized smooth muscle alpha actin antibody--a marker of myofibroblasts and smooth muscle cells--to study lipocytes in situ in normal and fibrotic rat liver as well as during their 'activation' in culture. Dual immunofluorescence studies on tissue sections from normal liver identified lipocytes as perisinusoidal desmin-positive, smooth muscle alpha actin-negative cells. In bile duct obstructed fibrotic liver, desmin-positive cells were numerous in areas of fibrosis and these cells also exhibited smooth muscle alpha actin. In carbon tetrachloride-induced fibrosis, cells expressing both desmin and smooth muscle alpha actin were present in fibrotic bands and in regenerating nodules. These results suggested that lipocytes had acquired characteristics of myofibroblasts during liver injury. To further address this issue we examined lipocytes immediately after isolation and also in primary culture. In freshly isolated lipocytes from normal liver, smooth muscle alpha actin was absent. In contrast, freshly isolated lipocytes from CCl4-treated animals expressed this smooth muscle marker immediately after isolation. In primary culture on plastic, lipocytes from normal liver began to express smooth muscle alpha actin coincident with culture-induced activation; at 14 days, smooth muscle alpha actin was identified in all cells. Electron microscopy demonstrated a highly developed array of microfilament bundles characteristic of actin filaments. Immunoblot of culture-activated lipocytes using the smooth muscle alpha actin antibody demonstrated the expected 42 kD protein (corresponding to the molecular size of smooth muscle alpha actin). Although smooth muscle alpha actin was readily detectable in culture-activated cells, it was not expressed in cells in which a quiescent phenotype was preserved by maintenance in culture on a laminin-rich gel. These findings demonstrate that the acquisition by lipocytes of a smooth muscle marker accompanies their 'activation', and are consistent with the hypothesis that lipocytes transform to myofibroblasts during liver injury.     

Immunohistochemical differentiation of leiomyocellular tumors and tumors with myogenic differentiation

The expression of alpha smooth muscle actin, muscle specific actin, desmin, h-caldesmon, and calponin was studied immunohistochemically in the following soft tissue and bone tumours and tumour-like lesions: muscle fibromatosis, inflammatory pseudotumours, chondroblastoma, enchondroma, chondrosarcoma, fibrous dysplasia, ossifying myositis, osteoblastoma, convential osteosarcoma, leiomyoma and leiomyosarcoma. Tumours and tumour-like lesions with myofibroblastic cells, osteoblasts and chondroblasts frequently exhibited intensive immunoreactivity for the muscle markers, and therefore, some of them may occasionally be confused with leiomyoma and leiomyosarcoma. Calponin does not help to differentiate various mesenchymal tumours expressing muscle markers, because it also stains intensively myofibroblasts, osteoblasts and chondroblasts. We confirmed that h-caldesmon was expressed intensely in leiomyomas and leiomyosarcomas, and never in the other tumours examined, with the exception of three chondroblastomas. The results have shown that h-caldesmon is a rather specific and sensitive marker for smooth muscle tumours, but it can also stain some actin positive myochondroblasts. It is possible that the positivity of h-caldesmon in some chondroblastomas is due to their complete myogenic transdifferentiation, and so we use the term myochondroblasts and myochondrocytes for designation of such S-100 protein, actin, and h-caldesmon positive cells.     

Cytoskeletal features of immature pulmonary vascular smooth muscle cells: the influence of pulmonary hypertension on normal development

Using an immunohistochemical technique, the development of the cytoskeletal proteins desmin, vimentin, and actin (using alpha isotype and non-isotype specific antibodies) was assessed using a semi-quantitative grading system in the pulmonary vascular smooth muscle of nine normal pigs and 19 normal humans at different ages, and in 13 children with pulmonary hypertensive congenital heart disease. In the normal of both species, immunostaining for vimentin decreased after birth and then increased gradually while immunostaining for desmin and alpha actin increased steadily with age. In pulmonary hypertension, immunostaining for alpha actin and vimentin showed an accelerated increase at between 2 and 8 months. Also, the media showed regional differences in immunostaining which preceded the development of intimal proliferation. The inner media showed less immunoreactivity for all cytoskeletal proteins studied than did the outer media. Within areas of intimal proliferation many cells were immunonegative. These results suggest that the cytoskeletal features of medial smooth muscle cells are remodelled in the normal infant; that this process is altered from at least 2 months in the pulmonary hypertensive infant; and that the smooth muscle cells immediately beneath the internal elastic lamina are remodelled before migrating to form intimal proliferation. Changes in cytoskeletal composition can be related to the previously described postnatal maturation of pulmonary vascular smooth muscle cells.     

大鼠骨髓间充质干细胞体外向血管平滑肌样细胞的诱导分化

背景：有研究表明,血管外膜成纤维细胞被激活早期参与动脉粥样硬化及血管再狭窄的形成,另有研究表明,骨髓间充质干细胞有部分黏附分化参与血管重塑,实验拟从血管外膜成纤维细胞与骨髓间充质干细胞之间相互作用的角度,来探讨动脉硬化及血管损伤后再狭窄过程中的可能机制。 目的：观察骨髓间充质干细胞与血管外膜成纤维细胞直接接触培养后,骨髓间充质干细胞形态改变及表达血管平滑肌肌动蛋白的情况。 方法：将骨髓间充质干细胞(DAPI标记细胞核)与外膜成纤维细胞按一定的比例混合培养7 d,细胞单独培养作对照,显微镜下观察细胞形态变化,免疫荧光检测骨髓间充质干细胞平滑肌肌动蛋白表达的情况。 结果与结论：骨髓间充质干细胞与血管外膜成纤维细胞共培养后,可见骨髓间充质干细胞的细胞核蓝染(DAPI标记细胞核)、胞浆红染(平滑肌肌动蛋白阳性)的双标细胞出现,且随培养时间的延长,骨髓间充质干细胞的平滑肌肌动蛋白表达阳性率明显增高。结果可见与血管外膜成纤维细胞直接接触有诱导骨髓间充质干细胞向血管平滑肌样细胞分化的趋势。     

Criteria for malignancy in nonvisceral smooth muscle tumors

Reliable prediction of the behavior of smooth muscle tumors is notoriously problematic at many anatomic locations. Relatively few studies have addressed the minimal histologic criteria for malignancy in smooth muscle tumors of soft tissue and the presently available criteria are not only inconsistently applied but also vary at different anatomic sites. It has been widely believed that the presence of any mitotic activity in a deep-seated smooth muscle tumor should lead to a diagnosis of leiomyosarcoma. Until very recently, the existence of leiomyomas in deep soft tissue has been disputed. Using presently available data, this review examines criteria for malignancy in smooth muscle tumors in each of the following anatomic compartments: deep soft tissue, including retroperitoneum/abdominal cavity; skin; subcutis; and external genitalia. Distinct criteria for assessing malignancy should be applied to smooth muscle tumors at each of these locations.     

Cytokeratins, smooth muscle actin and vimentin in human normal salivary gland and pleomorphic adenomas. Immunohistochemical studies with particular reference to myoepithelial and basal cells.

The distribution of immunostaining in normal major salivary gland and in 12 pleomorphic adenomas was studied using monospecific monoclonal antibodies to a number of cytokeratins, including cytokeratin 14, to smooth muscle actin and vimentin. A number of these antibodies enabled a distinction to be made between structural components of the normal gland, and to relate this to the different structures of pleomorphic adenomas. In the normal gland, the luminal duct cells expressed cytokeratins 7, 8, 18 and 19. Three antibodies were of particular value for the characterization of normal myoepithelial and basal cells; while the antibody to smooth muscle actin and the cytokeratin antibody Ks8.12 mutually exclusively stained the myoepithelial (basket) cells and the basal duct (light) cells, respectively, the recently established monospecific antibodies to cytokeratin 14 showed specific immunostaining with both cell types. These three antibodies left luminal cells virtually unstained. Ck 13 was found occasionally in single luminal excretory duct cells. Antibodies to cytokeratins 1/2, 10 and 10/11 did not show any staining in the normal gland. In the pleomorphic adenomas, the staining pattern of the two-layered tubular formation resembled that of the normal gland ducts: tumour luminal cells showed the characteristic, although more irregular, expression of cytokeratins 7, 8, 18 and 19; the outer cells resembled normal ductal basal cells with their anti-cytokeratin 14/Ks8.12-epitope staining and in that they virtually lacked staining for smooth muscle actin. Trabecular formations and cells in myxoid areas were reactive with Ks8.12 and for cytokeratin 14, occasionally also for cytokeratins 7, 18 and 19. Epidermoid cell islets expressed mainly cytokeratin 14 and inconsistently the squamous epithelial cytokeratin 13 and the epidermal cytokeratin 10/11. Vimentin was found in cells of myxoid areas. The results support the postulate that some of the normal duct basal cells act as reserve cells and can give rise to tumour formation with a primitive myxoid or trabecular pattern and a more differentiated tubular or epidermoid configuration.     

Expression of cyclooxygenase-2, c-kit, progesterone and estrogen receptors in uterine smooth muscle tumors: differential diagnosis

We examined the expression pattern of cyclooxygenase-2 (COX-2) and c-kit in uterine smooth muscle neoplasms and tried to determine the role of these markers in differential diagnosis. Archival tissue from 64 patients with uterine smooth muscle neoplasms (20 leiomyomas (LMs), 22 atypical leiomyomas (ALMs), and 22 leiomyosarcomas (LMSs)) was immunostained with antibodies against estrogen (ER) and progesterone receptors (PR), COX-2 and c-kit. 7 of 20 LM cases and 5 of 22 ALM cases were immunopositive for COX-2, whereas none of the LMS cases stained immunopositive (p< or =0.05). 4 of 20 LM cases and 5 of 22 ALM cases were immunopositive for c-kit, whereas 15 of 22 LMS cases showed c-kit immunopositivity (p< or =0.05). In conclusion, very few LMs and ALMs show COX-2 immunopositivity. LMSs usually do not express COX-2. COX-2 expression in smooth muscle tumors is not a prominent feature. Therefore, COX-2 inhibitors may not be useful in LMS therapy. C-kit was significantly expressed in uterine LMSs.     

Value of counting positive PHH3 cells in the diagnosis of uterine smooth muscle tumors

The diagnosis of uterine smooth muscle tumors including leiomyosarcomas (LMS), smooth muscle tumors of uncertain malignant potential (STUMP), bizarre (atypical) leiomyoma (BLM), mitotically active leiomyoma (MAL) and leiomyoma (LM) depends on a combination of microscopic features, such as mitoses, cytologic atypia, and coagulative tumor cell necrosis. However, a small number of these tumors still pose difficult diagnostic challenges. The assessment of accurate mitotic figures (MF) is one of the major parameters in the proper classification of uterine smooth muscle tumors. This assessment can be hampered by the presence of increased number of apoptotic bodies or pyknotic nuclei, which frequently mimic mitoses. Phospho-histone H3 (PHH3) is a recently described immunomarker specific for cells undergoing mitoses. In our study, we collected 132 cases of uterine smooth muscle tumors, including 26 LMSs, 16 STUMPs, 30 BLMs, 30 MALs and 30 LMs. We used mitosis specific marker PHH3 to count mitotic indexes (MI) of uterine smooth muscle tumors and compared with the mitotic indexes of hematoxylin and eosin (H&E). There is a positive correlation with the number of mitotic figures in H&E-stained sections and PHH3-stained sections (r=0.944, P<0.05). The ratio of PHH3-MI to H&E-MI has no statistically significant difference in each group except for LMs (P>0.05). The counting value of PHH3 in LMSs have significantly higher than STUMPs, BLMs, MALs and LMs (P<0.001) and the counting value of PHH3 is 1.5±0.5 times of the number of mitotic indexes in H&E. To conclude, our results show that counting PHH3 is a useful index in the diagnosis of uterine smooth muscle tumors and it can provide a more accurate index instead of the time-honored mitotic figure counts at a certain ratio.     

Mixed tumors of the vagina: an immunohistochemical study of 13 cases with emphasis on the cell of origin and potential aid in differential diagnosis.

Mixed tumors of the vagina (MTsV) are rare benign neoplasms characterized by an admixture of well-differentiated epithelial cells and stromal-type cells in various proportions. In contrast to mixed tumors in other anatomic sites, the histogenesis of the vaginal tumors is unclear. We studied the immunohistochemical profile of 13 examples to explore their histogenesis and determine whether their immunohistochemical profile might be useful in the differential diagnosis. The panel of antibodies used and the number of cases studied were: AE1/3 (12), cytokeratin 7 (CK7) (13), cytokeratin 20 (CK20) (13), epithelial membrane antigen (EMA) (13), muscle actin (MA) (12), desmin (11), h-Caldesmon (13), CD10 (13), CD34 (11), CD99 (8), and S-100 (7). Eight out of 12 tumors were positive for AE1/3, 7/13 for CK7, 2/13 for CK20, and 6/13 for EMA. MA was positive in 11/12 mixed tumors, desmin in 10/11 tumors and h-Caldesmon in 5/13. All tumors were extensively positive for CD10; CD34 was positive in 7/11; and none out of eight tumors showed membranous CD99 staining. Focal S-100 immunoreactivity was seen in 1/7 tumors. These results show that MTsV coexpress epithelial and mesenchymal markers. The expression of muscle actin (usually extensive), and focal desmin and h-Caldesmon positivity suggests the presence of a smooth muscle or myoepithelial component; however, the S-100 negativity and diffuse CD10 expression argue against it. Positivity for muscle markers does not help distinguish MTsV from smooth muscle or skeletal muscle tumors. The frequent expression of CD10 negates its use in the differential diagnosis with endometrial stromal tumors, and the CD10 and CD34 expression suggests that mixed tumors may arise from a primitive pluripotential cell. MTsV are positive for h-Caldesmon and CD10, two markers that have been used in gynecologic pathology primarily to aid in establishing the smooth muscle or endometrial stromal phenotype of a neoplasm.     

Characterization of natural Epstein-Barr virus infection and replication in smooth muscle cells from a leiomyosarcoma

Cells from a leiomyosarcoma tumor (LMS-1) from a patient with the acquired immunodeficiency syndrome (AIDS) were explanted, cultured in vitro, and studied by phase-contrast microscopy for morphologic and growth characteristics, immunostaining for cell markers, EBER in situ hybridization and polymerase chain reaction for detection of Epstein-Barr virus (EBV), and immunostaining for expression of EBV antigens. The cells exhibited very slow growth in vitro, with unusual elliptical and spindle-shaped morphology and fragmentation of the cytoplasm into long, tapering, cytoplasmic processes. Greater than 90% of cells expressed diffuse distribution of the smooth muscle isoform of actin by immunoperoxidase staining. Approximately 25% of cells expressed very bright fluorescence by immunostaining of the smooth muscle isoforms of calponin and actin. The majority of cells demonstrated a weak signal for CD21; approximately 5–10% of cells showed a strong signal that was confined to cell surfaces. The cultured cells harbored EBV, and infectious EBV continued to be detected by polymerase chain reaction and virus culture through several passages in vitro. Several EBV antigens were expressed, including latent antigen EBNA-1, immediate-early antigen BZLF1, early antigen EA-D, and late antigens, including viral capsid antigen p160, gp125, and membrane antigen gp350. Human umbilical cord lymphocytes that were transformed with virus isolated from cultured cells yielded immortalized cell lines that expressed EBV antigens similar to other EBV-transformed lymphocyte cell lines. These results confirm that EBV is capable of lytic infection of smooth muscle cells with expression of a repertoire of latent and replicative viral products and production of infectious virus. EBV infection of smooth muscle cells may contribute to the oncogenesis of leiomyosarcomas. J. Med. Virol. 57:36–46, 1999. © 1999 Wiley-Liss, Inc.     

Expression of smooth muscle markers in so called malignant fibrous histiocytomas

AIMS: To obtain further information regarding the frequency and degree of positivity for smooth muscle markers in a large number of malignant fibrous histiocytomas (MFHs), as an aid to accurate diagnosis. METHOD: The immunohistochemical features of 100 MFHs were studied and the results were compared with those for 30 leiomyosarcomas. Eighteen cases of MFH with smooth muscle actin (SMA) positivity were examined ultrastructurally. RESULTS: Immunoreactivity for smooth muscle markers, such as desmin, SMA, muscle specific actin (MSA) and h-caldesmon (HCD), which is a specific marker for smooth muscle cells and their tumours, was found in 28, 30, 29, and 29 of 30 leiomyosarcomas. Immunoreactivity for desmin, SMA, MSA, and HCD was found in 17, 30, 14, and two of the MFHs. On electron microscopic examination, approximately half of the cases contained a varying proportion of myofibroblastic cells. The others had only fibroblastic or undifferentiated tumour cells. At least 30% of the cases were found to display features consistent with limited smooth muscle or myofibroblastic differentiation. CONCLUSION: A large subset of so called MFH in fact shows poorly differentiated smooth muscle or myofibroblastic features, and perhaps such tumours should be regarded as pleomorphic leiomyosarcomas and/or pleomorphic myofibroblastic sarcomas.     

Fine structure of cystosarcoma phyllodes with reference to smooth muscle tumors

The fine structure of the stomal component of cystosarcoma phyllodes was compared with that of intracanalicular fibroademona, leiomyoma and leiomyosarcoma. The stromal cells of fibroadenoma was intermediate between fibroblasts and smooth muscle cells and may be designated as myofibroblasts. The stromal cells of cystosarcoma phyllodes have fulfilled the ultrastructural criteria for smooth muscle cells. Therefore, cystosarcoma phyllodes should be regarded as smooth muscle tumors or those of myoepithelial origin.     

Intermediate filament proteins and actin isoforms as markers for soft-tissue tumor differentiation and origin. III. Hemangiopericytomas and glomus tumors.

Intermediate filament proteins and actin isoforms of a series of 12 malignant hemangiopericytomas and five glomus tumors were examined by light microscopy, transmission electron microscopy, two-dimensional gel electrophoresis (2D-GE), and by immunohistochemistry, the latter using monoclonal or affinity-purified polyclonal antibodies to desmin, vimentin, cytokeratins, alpha-smooth muscle, and alpha-sarcomeric actins. By light microscopy, all hemangiopericytomas disclosed a predominant vascular pattern with scant storiform, myxoid and spindle cell areas, and with variable degrees of perivascular fibrosis. By ultrastructure, smooth muscle differentiation was observed in each hemangiopericytoma. Immunohistochemically, neoplastic cells of hemangiopericytomas expressed vimentin as the sole intermediate filament protein and lacked alpha-smooth muscle or alpha-sarcomeric actins. 2D-GE revealed only beta and gamma actins, in proportions typical for fibroblastic tissues. Glomus tumors revealed vimentin and alpha-smooth muscle actin within glomus cells by immunohistochemical techniques and disclosed ultrastructurally distinct smooth muscle differentiation. Therefore hemangiopericytomas represent a distinct soft-tissue neoplasm with uniform morphologic, immunohistochemical, and biochemical features most likely related to glomus tumors, the former representing an aggressive and potentially malignant neoplasm of vascular smooth muscle cells and the latter a well-differentiated neoplasm of vascular smooth muscle cells. Because malignant hemangiopericytomas disclose smooth muscle differentiation by ultrastructure, but do not express alpha-smooth muscle actin, as normal pericytes and glomus cells, it is suggested that these neoplasms represent highly vascularized smooth muscle neoplasms, ie, poorly differentiated leiomyosarcomas derived from vascular smooth muscle cells or their equivalent, the pericytes, which have lost alpha-smooth muscle actin as a differentiation marker that is similar to many conventional poorly differentiated leiomyosarcomas.