Results
The histologic pattern of fish in the creek was identified in all 15 cases of cutaneous angiosarcoma and was absent in the other 15 malignancies examined in this study. This finding shows the potential for the fish-in-the-creek pattern to be used as an additional diagnostic tool for dermatopathologists.
Comment
Cutaneous angiosarcoma is a rare but aggressive malignancy that proves difficult to diagnose both clinically and histologically as well as to treat effectively.1,5-8 Our results indicate that fish in the creek may be a useful and salient histologic feature in cutaneous angiosarcoma. It is important to recognize, however, that this finding should not be the sole feature upon which a diagnosis of cutaneous angiosarcoma is made, as it requires corroboration with positivity of MYC and D2-40 as well as a high Ki-67 proliferation index (>20%).11,12 Finding a fish-in-the-creek pattern should prompt dermatopathologists to consider a diagnosis of cutaneous angiosarcoma in the appropriate clinical and histologic settings.
The chief limitation of this study was the small sample size, with only 15 cases of cutaneous angiosarcoma available in the last 20 years at our institution. The limited sample size did not allow us to make claims on sensitivity and specificity regarding this histologic feature; however, with a larger sample size, the true diagnostic potential could be elucidated. Although the pathologists were blinded to the original diagnoses as they examined it for fish in the creek, it is possible they were able to make the correct diagnosis based on other histopathologic clues and therefore were biased.
Although the fish-in-the-creek pattern is present in cutaneous angiosarcoma, there may be other mimickers to consider. Intraluminal papillary projections lined by endothelial cells may be sectioned in a manner imitating this finding.3 In such a case, these endothelial cells must be differentiated from the free-floating or tufted spindle cells in order to have a positive finding for fish in the creek. There can be confusion if the biopsy cuts through a section of spindled cells, resulting in difficulty differentiating cutaneous angiosarcoma from other spindle tumors such as spindle cell melanoma or spindle cell squamous cell carcinoma.6 In such cases, immunohistochemistry may be helpful, as spindle cell melanoma would stain positive for S100 and SOX10 and spindle cell squamous cell carcinoma would stain positive for p63 and cytokeratin.
Various treatment strategies for cutaneous angiosarcoma have been employed, with the majority still resulting in poor outcomes.2,4-6 The recommended treatment is radical surgical excision of the primary tumor with lymph node clearance if possible. Following excision, the patient should undergo high-dose, wide-field radiotherapy to the region.5,8 Cutaneous angiosarcomas also have the ability to spread extensively through the dermis and can result in subclinical or clinically obvious widespread disease with multifocal or satellite lesions present. Distant metastases occur most frequently in the cervical lymph nodes and lungs.7 In cases where the disease is too extensive for surgery, palliative radiation monotherapy can be used.5,6
As atypical vascular lesions are considered to be a precursor to cutaneous angiosarcoma, it is important to note that the fish-in-the-creek feature was absent in all 6 of the atypical vascular lesions observed in the study. The differentiation generally is made based on MYC, which is present in cutaneous angiosarcomas and absent in atypical vascular lesions.10 The feature of fish in the creek may now be an additional clue for dermatopathologists to differentiate between angiosarcomas and other similar-appearing tumors.
Conclusion
Our study aimed to highlight an important histologic feature of cutaneous angiosarcomas that can aid in the diagnosis of this deceptive malignancy. Our findings warrant further study of the fish-in-the-creek histologic pattern in a larger sample size to determine its success as a diagnostic tool for cutaneous angiosarcomas. As noted previously, tumor grade does not impact survival outcome, but small tumor size has been one of the only features found to result in a more favorable prognosis.1,6,8 Future studies to identify a correlation between the histologic finding of fish in the creek and disease outcome in cutaneous angiosarcoma may be helpful to determine if these histologic findings provide prognostic significance in cases of cutaneous angiosarcoma.