Page 3 - Microsoft Word - 6942_BPR_Epub
P. 3
CUAJ – CUA Best Practice Report Guo et al
BPL: AML management
significantly lower fat content, traditionally referred to as minimal fat AMLs, and may not have this
9
characteristic appearance on US. To improve the accuracy of US, adjunct methods such as doppler or
contrast-enhanced US have been investigated. However, even with the use of both adjuncts, Ascenti et
al. reported a diagnostic accuracy of 78% when compared to pathological diagnosis. 10
Unenhanced computed tomography (UECT) is sensitive to detecting macroscopic fat in renal lesions.
Although attenuation values of < 10 HU in ROI are most often used to confirm fat, some have advocated
11
for a lower cut off of -15 or -30 HU to increase specificity. Thinner slices have also been demonstrated
to detect intralesional fat in smaller AMLs, with 3mm to 5mm slices identifying the vast majority of
lesions. 12,13
While the majority of AMLs can be diagnosed with UECT, the majority of the patients worked
up for an undifferentiated renal mass will undergo a multiphasic contrast-enhanced CT (CECT). AMLs
generally demonstrate homogenous enhancement, delayed washout and high intrinsic attenuation. The
addition of contrast does not add significantly to the sensitivity of CT for the diagnosis of AML. Woo et
al. published a meta-analysis of 15 studies with 2258 patients demonstrating multiple feature analysis of
CECT finding similar sensitivity to UECT (78% vs 81%). 14
Similar to UECT, MRI is excellent at identifying intralesional fat and may be more sensitive.
Classically, fat appears hyperintense on T1 sequences and hypointense on T2 images. However,
hemorrhagic cysts can have a similar appearance, and, in these cases, chemical shift fat suppression
18
sequences may be useful. This has also been shown to help identify minimal fat AMLs. Song et al.
reviewed 98 pathologically confirmed minimal fat AMLs and found that 23% of them were identifiable
19
on MRI but not CT. However, there remained another 23% of histologically confirmed AML that were
not discernable on CT or MRI. Song also proposed radiologically based categories for AMLs. Those
with fat visible on CT were termed “fat-rich.” The remainder, which would have been traditionally
called minimal fat AMLs, further subdivided into “fat-poor” and “fat-invisible.” Fat-poor AMLs were
only identifiable with additional MRI imagining, while fat invisible AMLs remain inconclusive. 19
Although the diagnosis of AML depends on the identification of intra-tumoural fat, some rare
fat-containing tumours may be malignant. Wilms tumours, extremely rare in adults, should be
considered in pediatric populations. Liposarcomas are most often perirenal rather than developing from
15
the kidney, and usually demonstrate renal displacement. Rarely RCC may contain fat, especially large
16
ones that entrap perirenal or sinus fat, or have calcifications representing osseous metaplasia.
Epithelioid AMLs (EAMLs) are a rare variant of AMLs that are composed of epithelioid cells,
20
with an absence of adipocytes and abnormal vessels. While classified with classic AMLs, they can
demonstrate malignant behaviour. The majority of evidence we have regarding EAMLs is from case
reports, and between 18 – 49% of these have been estimated to be malignant. 21,22 Given the controversy
over their malignant potential, some have further subdivided these lesions into pure EAMLs and AMLs
with epithelioid components, with pure EAMLs more likely to be considered high risk for metastatic
spread. However, while EAMLs belong to the same pathological family as AMLs, they rarely resemble
classic AMLs radiologically. The lack of adipose tissue in these lesions, particularly pure EAMLs, result