One hundred and one cases of clinical prostatic carcinoma (PCa), primary site, were analysed to define the interrelationship between tumour angiogenesis, histological grade, and bone marrow metastasis. Tumour angiogenesis was determined by the blood capillary density ratio (BCDR; a/b), defined as the ratio between the area of the blood capillaries (a) and the area of the tumour (b). The BCDR was evaluated by a colour image analysis system employing a computerized morphometrical method. A total of 43 cases of PCa with bone marrow metastasis (stage D2) and 58 cases of PCa without metastasis (stage B, C) were utilized. The prostatic carcinomas were classified into three groups (low, intermediate, and high) using Gleason's grading system. The BCDR of the primary PCa with bone marrow metastasis was similar in each of the three histologically graded scores. On the other hand, in the cases of PCa without metastasis, the BCDR of high score PCa was higher than those of the low and intermediate score PCa (U-test; P < 0.001). The BCDR of the high score PCa without metastasis was similar to that of the PCa with bone marrow metastasis. The BCDR may provide help in predicting tumour progression with regard to bone marrow metastasis of PCa with low and intermediate Gleason's scores.

The immunohistochemical expression of CD34 (human hematopoietic stem cell and endothelial cell marker) and laminin were studied in chronic liver diseases and hepatocellular carcinoma (HCC) to elucidate whether their expression reflected phenotypic differences between non-cancerous sinusoids and sinusoid-like tumor vessels. In normal liver, hepatic sinusoids were always negative for CD34 and laminin. In chronic hepatitis and cirrhosis, the two antigens were sparsely expressed in capillarized sinusoids at periportal and perinodular area. In advanced HCC, CD34 was strongly and diffusely expressed by the endothelial lining of sinusoid-like tumor vessels. However, early-stage HCC showed a wide spectrum of CD34 expression from negative to focal and diffuse, strongly positive staining in sinusoid-like vessels. Laminin was strongly expressed in advanced HCC but not in early-stage HCC. The results indicate that the enhanced expression of CD34 by sinusoidal endothelial cells may reflect the phenotypic change of endothelial cells in chronic liver diseases and HCC, and that the expression may correlate with the processes of angiogenesis induced by hepatocarcinogenesis.

Forty-seven cases of giant cell tumor of bone were clinicopathologically reviewed to determine any useful prognostic factors. Disease recurred in 11 cases. Eight of these cases had initially been treated with intracapsular piecemeal excision and three cases had been treated with wide excision. Nine of the 11 cases were classified as Grade III, two cases as Grade II, and one case as Grade II + fracture according to Campanacci's radiographic grading system. Intracapsularly excised cases had a high recurrence rate (47.1%). Metastasis to the lung occurred in three cases, each of which had been classified as Grade III. Although the radiographic Grade did not correlate with the rate of lung metastasis or recurrence, cases that metastasized to the lung or recurred tended to be radiographically aggressive. Disease recurred in eight of 24 Grade III cases; but in only two of 12 Grade II cases, in one of five Grade II + fracture cases, and none of six Grade I cases. p53 was expressed by mononuclear stromal cells in six cases. Disease recurred in four and lung metastasis occurred in three of these cases. p53 Expression correlated with rates of lung metastasis and recurrence. It was concluded that cases in which p53 is expressed have a high potential for lung metastasis and recurrence.

New monoclonal anti-MyoD1 and anti-myogenin antibodies were evaluated immunohistochemically to determine whether they are useful in discriminating rhabdomyosarcoma (RMS) from other soft tissue tumors in routinely processed sections. Neither MyoD1 nor myogenin was expressed in normal, mature striated muscle. In RMS, nuclear expression of MyoD1 and myogenin was found in 82 and 80% of non-overlapping cases, respectively. MyoD1 was generally expressed in small, primitive tumor cells, and larger cells exhibiting morphological evidence of skeletal muscle differentiation failed to express positive nuclear immunostaining. Positive nuclear staining for myogenin was stronger than that for MyoD1 in cases with abundant differentiated tumor cells, but was less prominent in cases in which small, primitive tumor cells predominated. No leiomyosarcomas, Ewing's sarcomas/peripheral primitive neuroectodermal tumors or other soft tissue tumors exhibited nuclear expression of MyoD1 or myogenin. In conclusion, both anti-MyoD1 and anti-myogenin antibodies are useful for diagnosing RMS and for discriminating RMS from other soft tissue tumors.

Our results suggest that tumour cells differentiating to Schwann cells are not the only component of MPNSTs. Furthermore, tumour cells other than Schwann cells are largely related to the formation of the heterogeneous components in MPNSTs associated with NF-1.