Fat Poor Renal Angiomyolipoma: Patient, Computerized Tomography and Histological Findings.

Purpose: We reviewed our experience with fat poor cases of angiomyolipoma. Materials and Methods: The records of patients with angiomyolipoma, as determined by pathological study, from 1998 to 2004 were reviewed by recording patient demographics and outcomes. Fat poor cases were defined as the failure of imaging to demonstrate fat in a lesion. Computerized tomography and histological characteristics were assessed. Results: Histologically confirmed angiomyolipoma was found in 15 patients. Multiple lesions were found in 3 of 15 cases (20%). Of these 15 patients who underwent surgery 11 (73%) had unsuspected angiomyolipoma due to absent fat on computerized tomography and they underwent intervention for presumed renal cell carcinoma. Mean age ± SD in this group was 54 ± 15 years and 8 of 11 patients (73%) were female, of whom 4 (50%) had uterine fibroids. These lesions were found incidentally in 7 of 11 cases (64%). Operative complications developed in 2 of 11 patients (18%). Average maximal diameter on pathological evaluation was 3.2 ± 1.3 cm (range 1.5 to 6). Nonenhanced computerized tomography was available in 7 of 11 cases, of which 3 of 7 (42%) showed hyperdense lesions and 4 of 7 (57%) showed isodense lesions. The percent of fat identified per high power field was less than 25% in 12 of 13 fat poor angiomyolipoma lesions (92%) compared to 2 of 4 classic lesions (50%) known to be angiomyolipoma before surgery (p = 0.04). Conclusions: We suggest that a general definition of fat poor angiomyolipoma should be the failure of imaging to reveal fat within a lesion, thus, making it unsuspected at surgery. A pathological definition should be less than 25% fat per high power field, which to our knowledge is a formerly undefined quantity. Not all cases are hyperdense on nonenhanced computerized tomography. These lesions cannot be reliably identified by imaging and they should be managed like all enhancing renal masses. [Copyright &y& Elsevier]