Leiomyoblastomas are rare tumors, and there have been few reports on their fine-needle aspiration (FNA) cytologic morphology. We herein describe the FNA features of a gastric leiomyoblastoma with subcutaneous metastasis. The cells had eosinophilic cytoplasm and oval eccentric nuclei, occasionally with intranuclear cytoplasmic inclusions. An organoid pattern was obvious in cell block sections. The tumor was positive for vimentin but negative for desmin, S-100 protein, and the common epithelial markers. The histogenesis is discussed, as are reasons why this tumor is better termed epithelioid mesenchymal tumor.

Background: Routine axillary dissection in breast cancer carries morbidity. Sentinel lymph node biopsy has been shown to predict axillary status. The experience of SLN biopsy in QEH is reviewed. Materials and methods: We included all patients with unifocal, palpable T1 or T2 tumour without palpable axillary LN from January 2001 to January 2004. Those who had previous excisional biopsy were also included. We excluded patients who had previous neoadjuvant chemotherapy or previous axillary surgery. Same‐day subdermal injection of C1Tc99m sulphur colloid and peritumoral patent blue injection technique was employed. SLN detected were removed, followed by formal axillary clearance. Outcome: The endpoint is to evaluate pathological result of SLN and remaining axillary content. Results: Forty‐three patients had SLN biopsy. Mean age was 51.2. Mean tumour size was 2.2 cm. Dye identification rate was 86%. Isotope identification rate was 97.7%. Mean no. of SLN harvested was 1.3. Sensitivity of SLN biopsy is 95.2%. Specificity is 100%. False negative rate is 4.76%. Negative predictive value is 95.7%. Accuracy is 97.7%. 57% of axilla‐involved patients had SLN as the only LN involved. Five out of 21 SLN metastasis were micrometastasis. 60% of micrometastasis were diagnosed by immunohistochemistry. Conclusion: Results of breast SLN biopsy in QEH is comparable to other centres and is feasible.

Backgrounds: Axillary lymph node(ALN) status has been important factor of treatment and prognosis for patients with breast cancer. Even though the better ultrasonographic instruments have been developed, it is still difficult to predict axillary lymh node metastasis (ALNM) with only ultrasonography(US) in T1 breast cancers which most of newly diagnosed breast cancers are recently since T1 breast cancers have low rate and less tumor burden of ALNM. This study evaluated the accuracy of prediction of ALNM in T1 breast cancer with US, contrast-enhanced MRI (cMRI) and contrast-enhanced 18F-FDG PET/CT (cPET/CT) and found out adequate combinations of these modalities. Method: Retrospectively, we reviewed 351 breast cancer patients with tumors(T1) ≤2cm in size between January 2008 and December 2011 who were preoperatively examined with US, cMRI, and cPET/CT and underwent pathologic evaluation of axillary lymph nodes acquired by sentinel lymph node biopsy or axillary dissection. Results: 94(26.8%) patients of 351 had ALNM. The sensitivity, specificity, positive predictive value(PPV), negative predictive value (NPV), and accuracy of ALNM with US were 0.457, 0.887, 0.597, 0.817, 0.772, respectively. cMRI had similar results with US. The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy of ALNM with cPET/CT were 0.447, 0.942, 0.737, 0.823, 0.809, respectively. The sensitivity if any one or more modalities were suspicious was 0.563. The specificity if all modalities were suspicious was 0.992. The PPV if cMRI and cPET/CT were suspicious was highest than if other combinations were suspicious. Conclusion: US, cMRI, and cPET/CT are helpful in prediction of ALNM of T1 breast cancers. However, there are no definite modality and combination of modalites to predict ALNM of T1 breast cancers. Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr P1-01-24.