Adrenal metastases: early biphasic contrast-enhanced CT findings with emphasis on differentiation from lipid-poor adrenal adenomas
F. Zhu, X. Zhu, H. Shi, C. Liu, Z. Xu, M. Shao, F. Tian J. Wang
Clinical Radiology 76 (2021) 294-301
Dr Rajiv B. Karia (Consultant GI/HPB and Interventional Radiologist, Chesterfield Royal Hospital & Nottingham University Hospitals, UK)
The adrenal gland is a site of both primary and metastatic malignant tumours as well as the most commonly identified benign tumour, adrenal adenomas. With a significant increase in use of CT for patients with previously known malignancies as well as for non-oncology patients, there has been an increase in number of incidentally identified adrenal tumours.
Adrenal metastasis (AM) and adrenal adenomas (AA), are the most common malignant and benign tumours[1]. Although on un-enhanced CT, attenuation of less that 10HU would be highly suggestive of an AA [1], approximately 30% of AAs are lipid poor[2] having a HU above 10[3] and hence un-enhanced CT are unable to reliably differentiate the two on un-enhanced CT.
Further imaging usually requires a chemical shift MRI or a delayed contrast-enhanced CT identifying relative washout of the lesion[4]. This retrospective study was used to identify imaging features to help differentiate AMs from AA, using un-enhanced CT and early biphasic contrast-enhanced CTs. This study identified 123 patients mostly based on histological diagnosis either at surgery or biopsy. However it also identified a few cases from follow-up imaging based on rapid increase in size (more than double within 6 months suggestive of AMs). One can be confident of the end point differential between AM and AA from this. All patients had unenhanced CTs, and biphasic arterial and portal venous CTs.
The study reviewed multiple imaging parameters; size: longest diameter (LD), shortest diameter (SD) and its ratio (LD/SD ratio); CT attenuation: unenhanced phase (CTu), arterial phase (CTa) and venous phase (CTv); degree of enhancement: degree of enhancement in arterial phase (DEAP), portal venous phase (DEPP); peak enhancement ratio PE/U (where PE was the highest value between DEAP and DEEP and U was the CTu). Further parameters such has absolute percentage washout (APW), and relative percentage washout (RPW) were also calculated.
This paper clearly identifies the presence of necrosis, location (unilateral/bilateral) and shape as significant differences between the two lesions. However, one can see the limitations of identifying necrosis particularly within smaller lesions, as it is at times difficult to differentiate necrosis from foci of low attenuation due to small quantities of microscopic fat component. This is not an issue with large lesions which are prone to cystic necrosis [5]. Although bilateral lesions maybe more commonly associated adrenal metastasis, we have all seen bilateral lipid poor AA, diagnosed on MRI. I think the qualitative assessment tools identified within this paper may not provide the reporter with the confidence to call AM over AA.
This paper has gone into great depth illustrating the quantitative analysis performed at identifying statistically significant radiological characteristics. The study identified as CTu, DEAP, DEPP, DEpeak and PE/U as all parameters providing statistically significance in differentiating AM from AA. However for a reporter to benefit from this, these findings need to simple to identify, a logical part of the reporting process and intuitive to use.
Univariate and ROC analysis showed PE/U ratio (<1.25), CTu (>32.2 HU), DEpeak (<43.15 HU), DEPP (<37.65 HU), presence of intralesional necrosis, location (bilateral adrenal), and irregular shape were independent factors for distinguishing AMs from AAs. However interestingly and more important for a reporter, when at least four of the criteria were combined, a sensitivity of 88% and specificity of 93% was achieved. I think with such high specificity using these parameters may prevent patients undergoing unnecessarily delayed ‘washout’ phase CTs or further assessment with time consuming MRIs.
The results of this study are worth noting; although this was a retrospective study, histological evidence was present in most cases; and there has been a great in-depth analysis of the parameters captured. There will obviously be some observer variation however two abdominal radiologists reviewed all the CT findings, blinded from the pathologies.
Often the findings of an incidental adrenal lesion, in day-to-day practice usually involve a single phase portal venous phase abdomen, and hence the data of the arterial, portal venous, and un-enhanced images are not always available. If this is the case, to further assess adrenal lesions, should one carry out an early biphasic/multiphase CT to capture the above, a delayed CT identifying lesional washout or an adrenal MRI? I suspect many places would still benefit from carrying out their usual practice and therefore I feel this will have very limited impact on the day-to-day practice.
However the paper has prompted discussion, and if we have these un-enhanced and bi-phasic data are we, as radiologists, confident enough with the above findings based on the 3 of the criteria’s to call an AM from an AA without the need for further imaging?
References:
- Meng X, Chen X, Shen Y, et al. Proton-density fat fraction measurement: a viable quantitative biomarker for differentiating adrenal adenomas from non-adenomas. Eur J Radiol 2017;86:112e8.
- Ganeshan D, Bhosale P, Kundra V. Current update on cytogenetics, taxonomy, diagnosis, and management of adrenocortical carcinoma: what radiologists should know. AJR Am J Roentgenol 2012;199(6):1283e93.
- Kim YK, Park BK, Kim CK, et al. Adenoma characterization: adrenal protocol with dual- energy CT. Radiology 2013;267(1):155e63.
- Park SW, Kim TN, Yoon JH, et al. The washout rate on the delayed CT image as a diagnostic tool for adrenal adenoma verified by pathology: a multicenter study. Int Urol Nephrol 2012;44(5):1397e402.
- Ctvrtlik F, Tudos Z, Szasz P, et al. Characteristic CT features of phaeo- chromocytomasdprobability model calculation tool based on a multicentric study. Biomed Pap Med Fac Univ Palacky Olomouc Czech Repub 2019;163(3):212e9.
Dr Rajiv Karia is a radiologist from Chesterfield Royal Hospital (Chesterfield, UK), specialising in GI/HPB and interventional radiology. Graduated from University of Nottingham and completed radiology training in Nottingham, having previously worked in surgery. Active ESGAR member since 2012, regularly attending annual meetings and has also completed the ESGAR/ESOR Exchange Programme for Abdominal Radiology Fellowship. Rajiv has a keen interest in technology and advances in medical applications, and is an associate member of the institution of Engineering and Technology (IET).
Comments may be sent to