Ultrasound liver imaging reporting and data system^®: Considerations for routine practice

Abstract

Hepatocellular carcinoma (HCC) is a growing problem worldwide. In Australia, both the incidence and mortality of HCC have sharply increased over the last few decades. Unless detected in its early stage, prognosis of HCC is grim. Ultrasound represents the main screening tool for surveillance programmes worldwide and is recommended in the guidelines by all relevant bodies. The American College of Radiology introduced the first edition of Ultrasound Liver Imaging Reporting and Data System (US LI‐RADS) in 2017 to propose an algorithm for reporting ultrasound studies performed in the setting of HCC screening and to provide technical recommendations. The algorithm has two components: detection scores and visualisation scores. More recently at our institution, we introduced a visualisation score to our routine work to assess visualisation of the liver. The score, which is applied by the sonographer performing the ultrasound study at the time of examination, comprises three categories that are not dissimilar to the score introduced by US LI‐RADS; adequate, adequate with limitations and inadequate. We believe that adopting a visualisation score is easy and important. It provides clinicians with information regarding the quality of the ultrasound study and the confidence in excluding HCC based on ultrasound alone. Attempts to improve ultrasound screening efficiency such as applying targeted liver ultrasounds and recording of a visualisation score are encouraged with the aim of improving patients’ outcomes.

Hepatocellular carcinoma (HCC) is a growing problem worldwide. It is the second most common cause for cancer‐related mortality and represents the fifth most common cancer in men and ninth most common cancer in women.1 In Australia, both the incidence and mortality of HCC have sharply increased over the last few decades. From 1982 to 2017, the age‐standardised incidence increased from 1.8 to 7.5 per 100,000 persons and mortality increased from 2.3 to 6.8 per 100,000 persons.2 These numbers have increased despite application of surveillance programmes and advancement in loco‐regional therapy options.

Unless detected in its early stage, prognosis of HCC is grim. Therefore, surveillance is advocated for patients with chronic liver disease at high risk of developing HCC. Ultrasound represents the main screening tool for surveillance programmes worldwide and is recommended in the guidelines by all relevant bodies.3 The use of additional screening tools such as tumour markers, namely alpha‐fetoprotein (AFP), is advocated to increase sensitivity. The combination of ultrasound and AFP on 6‐monthly basis is the current standard for surveillance in Australia. The guidelines recommend surveillance with ultrasound to be performed in a dedicated setting with experienced sonographers and established recall process to investigate findings detected on ultrasound.3 The American College of Radiology (ACR) developed the Liver Imaging Reporting and Data System (LI‐RADS^®) as a reporting system for computed tomography (CT) and magnetic resonance imaging (MRI) for patients at risk of developing HCC. The ACR has also introduced the first edition of Ultrasound Liver Imaging Reporting and Data System (US LI‐RADS^®) in 2017 to propose an algorithm for reporting ultrasound studies performed in the setting of HCC screening and to provide technical recommendations.4 We will briefly discuss the algorithm components, with a focus on the visualisation score, and compare the algorithm with our institutional practice considering implications for routine ultrasound screening.

The US LI‐RADS algorithm has two components: detection scores and visualisation scores. The detection score categorises findings into Negative, Subthreshold [observation(s) <10 mm in diameter but not definitely benign] and Positive [observation(s) ≥10 mm in diameter but not definitely benign OR new thrombus in vein]. These categories are not different from the routine findings documented by sonographers and reported for such ultrasound studies and are intuitive to every sonographer and radiologist. The visualisation score is an interesting component of this algorithm which is variably applied in daily practice and is not routinely reported. The score categorises ultrasound studies into A: minimal limitations, B: moderate limitations and C: severe limitations. The algorithm provides definitions for these visualisation scores with example descriptions. While this seems intuitive and known to the majority of sonographers and radiologists, formal inclusion of a score for liver visualisation in the ultrasound worksheet and report is not routine. From a practical perspective, most sonographers document the limitations of their studies when present such as body habitus and visualisation from intercostal approach only on the ultrasound worksheet but such limitations may not necessarily be reflected in the radiology report. Referring clinicians therefore may not necessarily be aware of such limitations which potentially decrease the sensitivity of ultrasound for detection of early HCC in some patients.

Compared to CT and MRI, ultrasound static images do not necessarily reflect the quality of the radiology examination. The performing sonographer, and sometimes radiologist, examines the area of interest and takes representative sample images and occasionally cine clips. The quality of the study can be judged by the reporting radiologist based on the images obtained. However, assessment of the quality of the study is often left to the performing sonographer and is therefore better documented on the ultrasound worksheet at the time of the study rather than retrospectively. Simmons et al.5 performed a retrospective analysis of 941 patients who had undergone a HCC screening ultrasound and divided them into four visualisation categories. The study was limited by the retrospective nature of static image quality assessment but concluded that 1 in 5 of these patients had an inadequate study.

At our institution, abdominal ultrasound studies performed for HCC screening are targeted to the liver, spleen and portal venous system and are tailored depending on the clinical indication. We routinely describe and report findings in relation to liver surface nodularity, echotexture, biliary tree and focal lesions as well as the portal venous system, spleen and presence of ascites. Tailoring the abdominal ultrasound reduces time spent on assessing other structures in the upper abdomen that are irrelevant to the clinical setting and improves efficiency.6 More recently at our institution, we introduced a visualisation score to our routine work to assess visualisation of the liver. The score, which is applied by the sonographer performing the ultrasound study at the time of examination, comprises three categories that are not dissimilar to the score introduced by US LI‐RADS; adequate, adequate with limitations and inadequate. Providing strict definitions for each category of this score was difficult, but a consensus amongst our group of sonographers and radiologists was reached. We defined an ‘adequate’ study as visualisation of both lobes of the liver in both subcostal and intercostal approaches (Figure 1a). An ‘adequate study with limitations’ was defined as visualisation of at least two‐thirds of the liver with suboptimal visualisation due to variety of reasons as listed in the inadequate category (Figure 1b). An ‘inadequate’ study was defined as visualisation of less than two‐thirds of the liver due to limitations such as very heterogeneous and/or fatty echotexture, high liver, large body habitus, chest wall deformity and/or bowel gas (Figure 1c).

Figure 1.

Adequate study (a), adequate study with limitations due to heterogenous echotexture (b) and inadequate study due to high liver position leading to poor visualisation (c).

We accumulated example ultrasound studies for each score from our HCC screening ultrasound patients’ cohort. The ultrasound studies served as a training set for all of our sonographers prior to rolling this process over into routine work. The ultrasound worksheet was amended to reflect this addition with tick boxes for each of the three categories for the visualisation score. We encouraged all our radiologists and registrars to include the adequacy of liver visualisation in their reports. Our institutional visualisation score is prospectively assigned by the sonographer performing the examination. This may not be the case with the US LI‐RADS visualisation score which might be assigned by the radiologist reporting the obtained static images although this issue has not been clearly stated in the US LI‐RADS document. There is currently no evidence at this stage to suggest that our approach is superior. However, it is well recognised that the quality of ultrasound findings is dependent on the skills and experience of the operator. Therefore, we believe that such visualisation scoring is better assigned by the operator.

Adopting the detection score of US LI‐RADS by identifying ultrasound studies as Negative, Subthreshold or Positive is not difficult to incorporate into routine work in a prospective manner or assess retrospectively from documented findings. We believe that adopting a visualisation score is easy and important. It provides clinicians with information regarding the quality of the ultrasound study and the confidence in excluding HCC based on ultrasound alone. Some patients will have inadequate ultrasounds, but the percentage and characteristics of such patients are yet to be determined because no prospective studies have been published on this topic and the US LI‐RADS has only recently been introduced. Such patients with inadequate ultrasounds unfortunately do not have a different screening pathway at this stage according to current guidelines.3 Identifying such patients as well as predictors for inadequate studies will provide information for future research into alternative or supplemental screening tools which could potentially be used in a risk‐stratified surveillance programme. At our institution, such patients are judged by referring specialists on a case‐by‐case basis and screening with CT or MRI is sometimes considered.

In summary, ultrasound remains the modality of choice for HCC screening and is likely to continue as such in the future given its worldwide availability and cost‐effectiveness as well as the absence of reliable biomarkers. Attempts to improve ultrasound screening efficiency such as applying targeted liver ultrasounds and recording of a visualisation score are encouraged with the aim of improving patients’ outcomes.

Disclosure statement

We declare that we have no conflicts of interest and that we received no financial support for this work.

Authorship declaration

We acknowledge that the authorship listing conforms to the journal's authorship policy and that all authors are in agreement with the content of the submitted manuscript.