1. Introduction
Renal cell carcinoma (RCC) is the most common malignant tumor involving the genitourinary tract, accounting for 2–3% of all human cancers [
1]. Incidence of renal cancer is highest in developed countries with rates generally higher in Europe and North America [
2]. Nowadays, most renal tumors encountered in clinical practice are small incidentalomas (cT1a) [
3].
Over the last few decades, the widespread use of abdominal imaging has increased the detection of localized, asymptomatic tumors, while less than 10% of patients currently present with the “classic triad” of symptoms: hematuria, flank pain and palpable masses [
4]. Nowadays, new diagnostic perspectives are offered by multiparametric ultrasound (US), in which US B-mode is combined with color Doppler investigation (CDI), contrast-enhanced ultrasound (CEUS) and elastography. Specifically, CEUS, allows a dynamic functional imaging of tumor perfusion [
5].
The benefit of using microbubble contrast agents in renal disorders has been well documented and the use of CEUS for renal indications has subsequently become part of the European Federation of Societies for Ultrasound in Medicine and Biology (EFSUMB) guidelines in 2008 with the latest update in 2017 [
6]. All approved US contrast agents have no renal excretion and renal insufficiency is not a contraindication for their use [
7]. Interestingly, CEUS is useful in the diagnosis of kidney lesions and during the follow-up of non-surgical lesions, representing a cost-effective and non-invasive method, not requiring the use of ionizing radiation.
Moreover, previous studies showed that CEUS can be used to differentiate among lesions with an equivocal enhancement at CT or MRI [
8,
9,
10,
11].
Therefore, CEUS has a potential role in the evaluation of renal lesions that deserve further investigation. Aim of this study is to evaluate the diagnostic usefulness of qualitative and quantitative CEUS for renal masses characterization as confirmed by histological examination.
2. Materials and Methods
2.1. Study Design, Inclusion and Exclusion Criteria
Between December 2019 and September 2021, a prospective single-center study was conducted at our institution. A total of 110 consecutive patients were enrolled. Overall, 118 renal masses, previously confirmed with CT or MRI, were studied with CEUS. Informed consent for the contrast-enhanced study was obtained from each patient. The study protocol was approved by the local ethical committee and was in accordance with the Helsinki Declaration of 1975. Inclusion criteria were: (1) renal mass detected by standard US; and (2) histopathological confirmation used as reference standard. Exclusion criteria were: (1) patients with contraindication for a contrast-enhanced study; (2) patients diagnosed with simple cysts on US B-mode; and (3) patients with metastases detected on preoperative imaging.
2.2. Imaging Acquisition and Interpretation
All baseline US and contrast enhancement ultrasound (CEUS) studies have been performed, with the same high-end US equipment (Toshiba Alpio 800, Tokyo, Japan) using 3.5-5-5 MHz convex probe. CEUS examination was performed by an expert radiologist with more than 20 years’ experience in genito-urinary ultrasound imaging. A 1 mL microbubble second generation contrast agent (SonoVue, Bracco Imaging, Milan, Italy) was administered via intravenous injection, followed by 5–10 mL of saline solution 0.9%. Microbubble enhancement has been followed continuously for at least 2 min. According to EFSUMB guidelines, a real-time video clip was recorded for each patient for review and documentation. After the completion of the imaging investigation, patients underwent surgical procedure of nephron sparing surgery (NSS). Histopathological examination has been performed and used asa reference standard in this study.
2.3. Morphological Pattern Analysis with US B-Mode and Color Doppler
Preliminary conventional US and CDI were performed, describing the lesion site, dimension, margin, echo-pattern and vascularization. According to US patterns, we classified renal masses presenting with regular margins and homogeneous hyperechoic or hypoechoic echogenicity as potentially benign. Masses presenting with irregular margins, inhomogeneous hyperechoic or hypoechoic echogenicity and/or with mixed appearance were considered potentially malignant. CDI patterns were classified as follows: type 1 (absence of intra-lesion vascularization); type 2 (peripheral vascularization); and type 3 (peripheral and intra-lesion vascularization). Type 1 has been considered benign, type 2 and 3 have been considered potentially malignant.
2.4. Qualitative CEUS Analysis
Parameters evaluated were: (1) The enhancement of the lesion, classified as rapid wash-in or with synchronous/late wash-in compared to surrounding healthy renal parenchyma; (2) homogeneous or inhomogeneous wash-in; and (3) the presence or absence of pseudo-capsule, defined as rim-enhancement. Potentially malignant parameters considered were the presence of rapid and inhomogeneous wash-in and rapid wash-out, when compared to healthy renal parenchyma. Potentially benign parameters considered were wash-in homogeneous and synchronous/late, when compared to healthy renal parenchyma.
2.5. Quantitative CEUS Analysis
Prospectively, videos recorded during CEUS have been used for elaborating time/intensity curves (TIC). Regions of interest (ROI) have been placed inside the lesion and within healthy renal tissue. Quantitative parameters of TIC have been automatically calculated by the software integrated in the US equipment. Finally, four different parameters were computed: (1) Peak of the signal intensity (PI) defined as maximum intensity value in time-intensity curve; (2) time to peak (sec) (TP) defined as the time elapsed between the moment when contrast medium first reaches the lesion and the time of maximum signal intensity after contrast medium administration; (3) mean transit time (sec) (MTT), defined as the lesion’s enhancement duration; and (4) AUC (area under the curve).
2.6. Statistical Analysis
Continuous variables were expressed as median and interquartile range (IQ), and discrete variables as numbers and percentages. Statistical analyses were performed using Stata 15.0 calculating sensibility, specificity, positive predictive value (PPV), negative predictive value (NPV) and area under the curve (AUC) of every parameter with dichotomous outcome. Reliability of measured parameters on a continuous scale has been calculated using the analysis of the ROC curve, in which the AUC has been measured and the Youden test has been performed to find out the cut-off to maximize diagnostic accuracy. The match between curves has been accomplished using the Bonferroni test with a statistical significance level of less than 0.05.
4. Discussion
CEUS represents a promising tool in the characterization of renal masses. In this study, the potential of this technique in a consecutive series of patients with renal masses has been prospectively investigated.
Our analyses resulted in several noteworthy observations.
First, we identified 110 patients with a diagnosis of renal masses treated by NSS. Of those, 8 patients presented with multiple masses, in consequence a total of 118 lesions were analyzed. Overall, the median tumor size was 2.8 cm and median R.E.N.A.L. nephrometry score was 7. Tumor lesions were sided either right (
n = 58) or left (
n = 60) and located at the superior (
n = 27), middle (
n = 40) and inferior (
n = 51) third of the kidney. Overall, renal masses were more frequently malignant vs. benign 75% vs. 25%, respectively. These histopathological findings are in line with other published series [
12].
Second, we analyzed the diagnostic performance of US B-mode parameters (echogenicity, vascularization and tumor margin) to distinguish malignant vs. benign renal masses. Here, inhomogeneous echogenicity resulted in the best accuracy (76%) relative to either vascularization (64%) or tumor margin (46%). Moreover, echogenicity characteristics resulted in interestingly high PPV (88%), but low NPV (57%). Notably, differentiating between US, benign solid renal masses from RCC remains a challenge and echogenicity of these masses is heterogeneously described in the literature [
13].
Third, we analyzed the diagnostic performance of CEUS qualitative parameters (presence of pseudo-capsule, inhomogeneous wash-in, rapid wash-in and rapid wash-out) in distinguishing malignant vs. benign renal masses. Here, the presence of pseudo-capsule resulted in the best accuracy (86%) relative to either rapid wash-in (76%) or inhomogeneous wash-in (77%) or rapid wash-out (67%). Moreover, the presence of pseudo-capsule resulted in interestingly high PPV and NPV of 94% and 71%, respectively. These observations are thus in agreement with the findings of previous studies, where the presence of a perilesional rim of enhancement resulted useful especially when distinguishing RCC from angiomyolipomas [
14,
15]. Moreover, homogeneous hypo- or iso-enhancement and rapid wash-in were associated with benign features in other historical studies [
16,
17].
Fourth, we analyzed the diagnostic performance of CEUS quantitative parameters, such as PI, AUC, MTT and TP. Our analyses did not allow us to set an adequate cut-off for both MTT and TP, and for this reason they were not included. Conversely, both PI and AUC yielded an accuracy of 93% and 95%, respectively. Moreover, AUC resulted in interestingly high PPV and NPV of 97% and 97%, respectively. On the other hand, PI showed a similar PPV of 96%, but a lower NPV of 87%. As a consequence, higher PI and increased AUC were two key differentiating features in predicting malignancy in the present study. These results were previously confirmed by Dai et al., where PI values were significantly higher in malignant lesions [
18]. By contrast, Xue et al. reported that TP outperformed PI in detecting malignancy [
19].
Fifth, we combined parameters with the highest accuracy in US B-mode (echogenicity) and in CEUS (PI and AUC). This combined score yielded a PPV of 100 % and NPV of 56% with an accuracy of 81%. Finally, after removing echogenicity, the combination of PI and AUC yielded a PPV of 100% and NPV of 83% with an accuracy of 93%. Taken together, CEUS parameters showed high values of diagnostic accuracy. Conversely, B-mode US parameters yielded a low accuracy rate with the exception of echogenicity. In consequence, we recorded an additional benefit when adding CEUS parameters in the analyses of renal masses. These results are promising, since both PI and AUC are quantitative parameters and for this reason are widely reproducible. Here, we adopted cut-offs calculated in our preliminary analyses (Youden test). Nevertheless, these cut-offs need to be validated in an external cohort to test their reliability.
Our work in not devoid of limitations and should be interpreted in the context of its single-center cohort. First, the small sample size and relatively small percentage of benign tumors, only 25% of our sample. Second, we only enrolled patients with an indication for surgery, in consequence, we potentially excluded all those patients undergoing active surveillance from the study. Third, all cases were evaluated by a single reader with long experience in urological ultrasound imaging. For this reason, our results may overestimate the diagnostic accuracy of this technique and potentially undermine their reproducibility in clinical practice. Further studies are needed to evaluate the diagnostic accuracy of CEUS in characterizing renal masses and to assess inter-reader variability for sonologists with different expertise.