Leiomyosarcoma with partial rhabdomyoblastic differentiation: First case report of primary cardiac origin

Primary cardiac tumors represent a rare neoplastic condition with an incidence that ranges from 0.0017 to 0.019% [1], of which 25% are malignant. Among such tumors, angiosarcoma is the commonest malignant tumor followed by rhabdomyosarcoma, malignant mesothelioma, and fibrosarcoma, each with an incidence that is greater than 10% [2]. However, the incidence of cardiac leiomyosarcoma is less than 1% [2]. Previous studies of leiomyosarcoma with rhabdomyoblastic differentiation have conducted to those arisen from another site [3‐11], and they announced a poorer prognosis of this tumor. Especially, Oshiro et al. have reported that leiomyosarcoma with rhabdomyoblastic differentiation shows poorer prognosis than typical leiomyosarcoma [6]. In the present paper, we describe an extremely rare primary cardiac malignant tumor. To the best of our knowledge, this is the first case of primary cardiac leiomyosarcoma with partial rhabdomyoblastic differentiation.

Leiomyosarcoma occurring as a primary cardiac tumor has been known as an extremely rare condition of which the rate represents less than 1% of all primary cardiac malignant tumors [2]. Furthermore, to the best of our knowledge there has been no report of a case of primary cardiac leiomyosarcoma with partial rhabdomyoblastic differentiation. In general, leiomyosarcoma is currently subdivided histologically into four types: classical, epithelioid, pleomorphic, and myxoid [13]. The morphological findings of the present case indicated a myxoid type of leiomyosarcoma, but immunohistochemistry revealed that a few tumor cells showed positive reactivity for myogenin. This has been known as a myogenic transcriptional regulatory protein which is expressed in the early phase of skeletal muscle differentiation (rhabdomyogenic differentiation), and it induces differentiation of myoblasts into the multinucleated myotube [14]. This myogenic regulatory protein has been largely accepted as a sensitive and specific immunohistochemical marker for rhabdomyosarcoma or other tumors with rhabdomyoblastic differentiation [14].

Meanwhile, it is interesting to note that the spindle cell showed positive reactivity for CK CAM 5.2. Although, it has been well known that leiomyosarcoma usually showed negative reactivity for epithelial markers [15], some investigators described that a part of leiomyosarcoma shows positive reactivity for CK [15‐17]. Therefore, CK CAM 5.2 expression in the present case may support a diagnosis of leiomyosarcoma. However, to make diagnosis of leiomyosarcoma with rhabdomyoblastic differentiation, we should refer three important tumors and deny them, respectively, which comprise undifferentiated pleomorphic sarcoma (UPS), rhabdomyosarcoma, and rhabdomyoma. Cardiac UPS usually occurring at the left atrium, histopathologically comprises a mixture of spindle cells in a storiform pattern with polyhedral cells [5]. Furthermore, high-grade undifferentiated sarcomas can exhibit focal α-SMA expression [15]. These findings are similar to the present case. However, the spindle cell, a major component of the present tumor, had an elongated, blunt-ended, and hyperchromatic nucleus plus spindle, fibrillated, and eosinophilic cytoplasm. In addition, the cell showed positive reactivity both for α-SMA and desmin, focally, by immunohistochemical examination. These findings allowed disclosing the smooth muscle differentiation. Furthermore, some of the spindle cell also showed positive reactivity for CK CAM 5.2, of which positive ratio has been reported ranging from 22.2% (2/9) to 35.0% (14/40) in leiomyosarcoma [16, 17]. Although it still remains a difficulty for decision, we made the diagnosis of leiomyosarcoma rather than UPS. On the other hand, since rhabdomyosarcoma has been know as the second most common primary cardiac malignant tumor [2], that should also be considered as a disease for differential diagnosis. Especially, embryonal rhabdomyosarcoma usually shows similar morphologic findings of the present case, such as varying degrees of cellularity containing hypercellular and loosely textured myxoid areas, hyperchromatic and round or spindle-shaped nucleus, and eosinophilic cytoplasm [18]. However, embryonal rhabdomyosarcoma is uncommon in patients older than 40 years of age [18] and neither cross-striation nor myoglobin expression was proven in the present case. Furthermore, a large body of spindle cells showed negative reactivity for myogenin (only 2.1% of them showed positive reactivity) that has been largely accepted as a sensitive and specific immunohistochemical marker for rhabdomyosarcoma or other tumors with rhabdomyoblastic differentiation [14]. According to our immunohistochemical examinations, we were able to deny typical rhabdomyosarcoma. As for rhabdomyoma, the most common subtype of cardiac origin has been known as cardiac rhabdomyoma, but it occurs almost exclusively in the hearts of infants and young children and composes predominantly large polygonal vacuolated spider cells [19]. Therefore, the adult type of rhabdomyoma should be considered as differential diagnosis which is usually composed of tightly polygonal cells which had peripherally placed nucleus plus acidophilic, finely granular, and vacuolated cytoplasm. However, mitotic figures are nearly absent, cross-striations can be discerned, and show positive reactivity for rhabdomyogenic markers immunohistochemically in these two subtypes of rhabdomyoma [19]. These results were different from the findings extracted from the present case.

On the other hand, only one case of sarcoma arisen from myocardium with rhabdomyoblastic differentiation has been reported by Kabir et al. [20] who described a malignant peripheral nerve sheath tumor indicated an area of rhabdomyoblastic differentiation in part. In their report, a little information of immunohistochemical examinations was described which simply comprised positive reactivity for s-100 protein and focal for desmin. These results were different from these of the present case, but comparative discussion could not be completed in detail. Therefore, we preferred to diagnose this primary cardiac malignant tumor as a leiomyosarcoma with partial rhabdomyoblastic differentiation. Previous studies of leiomyosarcoma with rhabdomyoblastic differentiation have conducted to those arisen from another site, and Oshiro et al. have reported that leiomyosarcoma with rhabdomyoblastic differentiation shows poorer prognosis than typical leiomyosarcoma [6]. In fact, the present case showed rapid growth of the tumor and the patient died due to extensive metastases in the lungs despite early diagnosis and surgical removal. Table 2 presented herein summarizes major clinical data of nineteen cases of leiomyosarcoma in soft tissue with rhabdomyoblastic differentiation, and includes the present case representing the first report of leiomyosarcoma arising from the myocardium [table 2]. The patient age ranged from 33 to 85 (mean: 62.7). The male-to-female ratio was 10:10. The tumor sizes ranged from 20 to 250 mm (mean: 116.4).

We wish to take a more detailed discussion on the present case, especially in relation to tumor cell differentiation with phenotypical expression analysis. The spindle cell, a major component of the tumor, has the potential to differentiate into a smooth muscle cell which can be phenotypically identified with positive reactivity for α-SMA. However, we found 2.1% of spindle cells showed positive reactivity for myogenin, and half of myogenin-positive spindle cells showed positive reactivity for α-SMA at the same time. In contrast, none of polyhedral cells showed positive reactivity for α-SMA, but they exhibited a significantly higher myogenin-positive rate than spindle cells. These findings support the following hypothesis. First, the "rhabdomyoblastic differentiation" observed in the present case may represent the early stage of rhabdomyogenic differentiation because tumor cells showed positive reactivity for myogenin which has been known as a maker of cells in the early phase of rhabdomyogenic differentiation, and exhibited neither cross-striation nor myoglobin expression. Second, polyhedral cells may be in a more advanced stage of rhabdomyogenic differentiation than spindle cells because polyhedral cells were morphologically similar to rhabdomyoblasts and showed a significantly higher myogenin-positive rate than spindle cells.

Finally, the possibility of transdifferentiation or synchronous smooth and skeletal muscle differentiation in leiomyosarcoma was suggested in the present case because synchronous expression of α-SMA and myogenin was confirmed in 1.0% of spindle cells. Overall, our immunohistochemical evaluation indicated that rhabdomyoblastic differentiation in leiomyosarcoma might be generated not only by de novo generation from mesenchymal cells.

We describe an extremely rare case of primary cardiac leiomyosarcoma with partial rhabdomyoblastic differentiation. The tumor indicated aggressive growth and the patient died despite early diagnosis and surgical removal. Furthermore, our immunohistochemical evaluation suggested that rhabdomyoblastic differentiation in leiomyosarcoma might be generated not only by de novo generation from mesenchymal cells. To the best of our knowledge, this is the first report of primary cardiac leiomyosarcoma with partial rhabdomyoblastic differentiation.

We could not get the proof of the patient's written and signed consent for the publication because we could not announce disease of our patient to herself due to her family's request. Furthermore, the patient has already died. However her family agreed our proposal using surgical specimen for our research and written informed consent was obtained from the patient's family (as a proxy) for publication of this case report and any accompanying images. A copy of the written consent is available for review by the Editor-in-Chief of this journal.