Renal cancer accounts for 2% to 3% of all malignant diseases with an increasing incidence worldwide [1, 2]. The majority of renal cancers are asymptomatic and slowly growing small renal masses (SRMs), which are incidentally detected by the commonly used modern imaging procedures such as ultrasonography (US), cross-sectional computed tomography (CT), and magnetic resonance imaging (MRI) in the last three decades. SMRs are defined as predominantly solid-enhancing tumors measuring smaller than 4 cm in maximal diameter [3], representing both benign and malignant lesions. Studies have demonstrated an inverse correlation between lesion size and malignancy. Specifically, approximately 6% of masses larger than 7 cm were benign, 19% to 22% of masses between 1 and 4 cm were benign, and up to 46% of masses less than 1 cm were benign [4]. Most angiomyolipomas (AMLs) need only active surveillance rather than invasive treatment, while for renal cell carcinoma (RCC), especially for clear cell RCC, surgical resection is the preferred therapy. Although most of the SRMs are assumed to be early-stage disease, the mortality rate for RCC has remained unchanged during the past several years [5]. Therefore, it is essential to differentiate benign from malignant SRMs to better direct prognoses and management.

Renal biopsy is a useful method to distinguish benign from malignant renal masses preoperatively; however, as mass size decreases, the sensitivity of biopsy also falls [6]. In addition, the potential for sampling error and complications such as bleeding and tumor cell seeding lead to the low overall use of biopsy for SRMs. Therefore, non-invasive and reproducible methods for the efficient differential diagnosis of SRMs are needed.

Contrast-enhanced CT (CECT) is most often used for the assessment of renal masses. However, CECT has shortcomings, including exposure to ionizing radiation, as well as the use of potentially nephrotoxic and immunogenic iodinated contrast. The latter is especially problematic in patients who have some degree of chronic renal failure. Contrast-enhanced ultrasound (CEUS), which is a relatively novel modality with microbubble-based contrast agents, has played an important role in the diagnosis and differential diagnosis of benign and malignant renal masses in recent years [7]. As the size is similar to red cells (ranging from 1 to 10 μm) but much larger than particles used as CECT, the microbubbles remain completely intravascular and have no nephrotoxicity, thus enabling their safe use [8]. The lack of ionizing radiation, cost-effectiveness, non-invasive nature, and real-time imaging make CEUS very useful for the diagnosis of kidney lesions and follow-up imaging. The main limitations are that it is dependent on operator experience, image acquisition is influenced by the location of the lesion and the patient’s constitution, and the images are less reproducible and have a wide range of sensitivity and specificity [9].

A previous study showed that CEUS was better than CT in the diagnosis of malignancy in complex cystic renal masses [10]. To our knowledge, there is no study to date comparing CEUS and CECT in differentiating benign from malignant SRMs. We sought to compare the diagnostic efficiency of CEUS with that of CECT in the evaluation of benign and malignant SRMs that were confirmed by pathology retrospectively.

Materials and methods

Patients

This retrospective study was approved by the ethical committee of our hospital and informed consent was waived. From January 2013 to August 2016, 128 patients (79 males and 49 females, age range 26–78 years, mean age 53.5 ±12.6 years) with SRMs underwent both CEUS and CECT scans preoperatively at our institution. All were confirmed by histopathologic diagnosis after surgical resection. The patients with masses greater than 4 cm, with simple or complex cysts (n = 2), cystic tumors (n = 5), and with multiple or metastatic tumors (n = 3) were excluded from this study. Thus, a total of 118 patients (73 males and 45 females, age range 26–78 years, mean age 52.8 ±11.2 years) with 118 renal masses smaller than 4 cm were evaluated in this study, including 25 benign lesions and 93 malignant lesions, and their medical records and diagnostic imaging studies were retrospectively reviewed.

CEUS imaging technique

CEUS examinations were performed using a Sequoia 512 ultrasound system (Siemens, Mountain View, CA, USA) and a 4C1 probe with a frequency of 3–4 MHz. The contrast pulse sequence (CPS) imaging mode was used with a low mechanical index (MI) ranging from 0.18 to 0.20. Initial conventional ultrasound scanning was conducted to obtain the location, size, and internal echogenicity of the tumors using gray-scale imaging, and the blood flow of the tumor was assessed by color Doppler flow imaging (CDFI). The depth of the renal lesions was all less than 10 cm in this study. Optimal scanning containing both mass and normal adjacent renal parenchyma was selected, and the system was then switched to contrast imaging mode using contrast pulse sequencing (CPS) technology. The second-generation contrast agent SonoVue (Bracco SpA, Milan, Italy) was used in this study. The SonoVue freeze-dried powder, which contained sulfur hexafluoride-filled microbubbles, was dissolved in 5 ml of normal saline before injection, and a dose of 1.6–2.4 ml contrast agent, which was chosen individually depending on the weight, height, and age of the patient, was used with an intravenous injection in a bolus fashion, followed by 5 ml of saline flush. The timer and video recorder were started immediately after the injection of the contrast agent. All patients were required to maintain slow shallow breathing when CEUS was conducted, and the duration of the complete examination was at least 3 min. The contrast phase terminology of the renal CEUS examination is controversial with numerous naming schemes used in previous literatures. In this study, we used the following phase terms: cortical phases, began 10–15 s after injection until to 30–45 s, and medullary phases, approximately 30–45 s after injection until the microbubble echoes disappeared completely [11, 12]. If necessary, a second injection of the contrast agent was conducted 15 min after the first injection. The single images and video clips of the CEUS were digitally stored in the hard drive of the system for documentation and subsequent qualitative analysis. All the US scanning was performed by one investigator who had 8 years of experience in abdominal ultrasonography and CEUS studies and was blinded to the diagnosis of renal masses.

CECT imaging technique

All examinations were performed on a dual-source, dual-energy CT scanner (Somatom Definition, Siemens Medical Solutions, Germany). Both abdominal unenhanced CT and CECT scans were performed. Parameters included a detector collimation of 64 × 0.6 mm2, a pitch of 1.2, a gantry rotation time of 0.5 s, a tube voltage of 120 kVp, and an abdominal reference tube current of 210 mAs. Automated tube current modulation was used in all CT studies (CAREDose 4D; Siemens Medical Solutions). All images were reconstructed from the CECT scan with a 0.75-mm slice thickness and a 0.5-mm reconstruction increment. The CECT scan was started by continuously injecting a bolus of 80 ml of iopromide (Ultravist; 300 mg I/mL, Bayer Schering Pharma, Berlin, Germany) followed by 40 ml of saline solution into an antecubital vein via an 18-gauge catheter (injection rate: 5 ml/s). The enhanced CT scans were performed at a 25–30 s corticomedullary phase, 55–60 s nephrographic phase, and 240 s excretory phase (delayed phase).

Imaging interpretation

The CEUS images were reviewed by two independent ultrasonologists blinded to the clinical histories, CT findings, and final histological results in a random order. Both the ultrasonologists had more than 10 years of experience in CEUS evaluation and were not involved in the original ultrasonographic and CEUS scanning. The imaging characteristics included the mass position, size, echogenicity, homogeneity, the presence of a hypoechoic rim, and color flow signals on conventional and color Doppler ultrasound imaging. The “wash-in” and “wash-out” pattern, the degree of peak enhancement, the homogeneity of enhancement, and the rim enhancement were also evaluated by CEUS imaging. The normal renal cortex adjacent to the tumor was used as the control for comparison of the enhancement. The “wash-in” and “wash-out” pattern was classified as fast, simultaneous, or slow. The degree of tumor peak enhancement was classified as hypo-, iso-, and hyperenhancement [13]. The homogeneity at peak enhancement was classified as homogeneous, which was defined as a lesion completely enhancing without any defects, and heterogeneous, which was defined as a lesion with unenhanced areas, regardless of various enhancement degrees [14].The perilesional rim enhancement was considered to represent the presence of a pseudocapsule, which appears as a hypoechoic rim during the early corticomedullary phase, becoming hyperechoic and clearly visible during the late nephrographic phase [15].

Two radiologists (with 20 and more than 10 years of experience in CT examination of the urinary tract, respectively), blinded to the clinical histories, ultrasound findings, and final histological results retrospectively reviewed the CECT images independently. The imaging parameters included the mass position, size, margins, cystic components or necrosis, calcification, mass attenuation on unenhanced CT scan, the degree of enhancement (in Hounsfield units, HU) on different phases of the CECT scan, homogeneous or heterogeneous appearance, lymphadenopathy, perinephric stranding, hydronephrosis, the presence or absence of a clear capsule sign, and vascular invasion.

To evaluate whether the renal lesions were benign and malignant, all lesions on both CEUS and CECT were independently scored on a 3-point scale: a score of 1 indicated a benign lesion, a score of 2 indicated an equivocal lesion, and a score of 3 indicated a malignant lesion. Cases in which the readers disagreed were reviewed by both investigators to obtain a final consensus result. Table 1 shows the diagnostic criteria of CEUS and CECT according to the established criteria developed on the basis of enhancement patterns of renal masses previously described [16–21].

Table 1 Diagnostic criteria at CEUS and CECT
Full size table

Statistical analysis

All statistical analyses were performed using a statistical software package (MedCalc Software; MedCalc, Mariakerke, Belgium). Data were expressed as mean (standard deviation) or as median (interquartile range). Differences between groups were analyzed using Student’s t-test for continuous variables, and chi-square and Fisher’s exact tests for categorical variables. The sensitivity, specificity, positive predictive values (PPV), negative predictive values (NPV), and area under curve (AUC) were calculated with 95% confidence intervals (CI) to assess the performance of CEUS and CECT to differentiate benign from malignant lesions and were compared according to the gold standard (histopathology) using McNemar’s test and the Z test. A weighted kappa was used to interpret the concordance between interobserver agreement. Kappa values were interpreted as follows: 0–0.20, poor agreement; 0.21–0.40, fair agreement; 0.41–0.60, moderate; 0.61–0.80, good; and 0.81–1.00, very good agreement. A P value of less than 0.05 was considered statistically significant.

Results

Patients and tumor characteristics

The 118 included lesions had a mean size of 2.75±0.73cm (range 1–3.9 cm), all the lesions were detected by both CEUS and CECT, 25 cases were benign lesions (25/118, 21.2%), including 20 AMLs, 3 oncocytomas, and 2 metanephric adenomas (MAs); 93 cases were malignant lesions (93/118,78.8%), including 75 clear cell RCCs (ccRCCs), 13 papillary RCCs (pRCCs), 2 chromophobe RCCs (chRCCs), 1 unclassified RCC, 1 Xp11.2 translocation/TFE3 gene fusion RCC (Xp11.2/TFE3 RCC), and 1 mucinous tubular and spindle cell carcinoma (MTSCC). All diagnoses were confirmed by histopathology after surgery. Table 2 describes the patients and tumor characteristics between benign and malignant lesions. Patients with malignant lesions were more often male and those with benign lesions were more often female (P < 0.05). The benign lesions that were treated with nephron-sparing partial nephrectomy were more and those treated with radical nephrectomy were less than malignant lesions; there was significant difference in operation method between benign and malignant lesions (P < 0.05). There was no significant difference in age, size, tumor side, and location of tumors between benign and malignant lesions (P > 0.05).

Table 2 Patients and tumor characteristics between renal benign and malignant small renal masses
Full size table

Interobserver agreement

The CEUS features in benign and malignant SRMs for reader 1 and reader 2 are summarized in Supplement Table 1. In the CEUS imaging evaluation of the 118 lesions, the two readers were in agreement over the scores of 112 lesions, among which 22 were scored 1, and 90 were scored 3. The readers disagreed over the scores of six lesions: two were scored 1 by reader 1 and scored 3 by reader 2, one was scored 2 by reader 1 and scored 1 by reader 2, two were scored 2 by reader 1 and scored 3 by reader 2, and one was scored 3 by reader 1 and scored 1 by reader 2. All six lesions were reassessed, as 1 lesion was scored 1 and 5 lesions were scored 3 with a consensus. The weighted kappa value was 0.89 (95% CI 0.79–0.98). The CECT features in benign and malignant SRMs for reader 1 and reader 2 are summarized in Supplement Table 2. In CECT imaging evaluation, the two readers were in agreement over the scores of 113 lesions, among which 26 were scored 1, and 87 were scored 3. The readers disagreed over the scores of five lesions: one was scored 1 by reader 1 and scored 3 by reader 2, two were scored 2 by reader 1 and scored 1 by reader 2, and two were scored 3 by reader 1 and scored 2 by reader 2. All five lesions were reassessed, as 3 lesions were scored 1 and 2 lesions were scored 3 with a consensus. The weighted kappa value was 0.93 (95% CI 0.87–0.99) (Table 3).

Table 3 Concordance between interobserver agreement on the scores of small renal masses
Full size table

Comparison of qualitative diagnostic accuracy between CEUS and CECT

For benign lesions, in 20 AMLs (including 10 classic AMLs and 10 lipid poor AMLs), 15 cases (including 10 classic AMLs and 5 lipid poor AMLs) were demonstrated by both CEUS and CECT (Fig. 1). The other 5 lipid poor AMLs were misdiagnosed as malignant lesions (RCCs) by either or both methods, among which 2 were misdiagnosed by CEUS, 1 was misdiagnosed by CECT, and 2 were misdiagnosed by both CEUS and CECT (Fig. 2). The three oncocytomas were all misdiagnosed by both CEUS and CECT as malignant lesions (RCCs). In two MAs, one was demonstrated by both CEUS and CECT and the other was misdiagnosed by CEUS as a malignant lesion (RCC). For malignant lesions, in 75 ccRCCs, 66 were demonstrated by both CEUS and CECT (Fig. 3), 5 were misdiagnosed by CEUS as benign lesions (AMLs), 3 were misdiagnosed by CECT as benign lesions (AMLs), and 1 was misdiagnosed by both CEUS and CECT as a benign lesion (AML). In 13 pRCCs, 5 (5/13, 38%) cases demonstrated no or equivocal enhancement using CECT; thus, 8 were demonstrated and 5 were misdiagnosed by CECT as benign lesions (2 complex cysts and 3 AMLs) (Fig. 4), while all lesions demonstrated enhancement by CEUS. The 2 chRCCs, 1 Xp11.2 /TFE3 RCC, and 1 MTSCC were all demonstrated by both CEUS and CECT, while the one unclassified RCC was demonstrated by CEUS but misdiagnosed by CECT as a benign lesion (AML). To compare the qualitative diagnostic accuracy on different histotypes of SRMs between CEUS and CECT, the results are summarized in Table 4. As there were few cases of oncocytoma, MA, chRCC, unclassified RCC, Xp11.2/TFE3 RCC, and MTSCC in this group, the statistical analysis was carried out only on the diagnosis of AML, ccRCC, and pRCC. The qualitative diagnostic accuracy on pRCC by CEUS was better than that by CECT (P < 0.05), while there was no significant difference in qualitative diagnostic accuracy on other histotypes of renal lesions between CEUS and CECT (P > 0.05).

Fig. 1
figure 1

A 54-year-old woman with a 2.1-cm diameter AML was diagnosed using both CEUS and CECT. A Conventional ultrasound demonstrated a hyperechoic mass located in the interpolar pole of the left kidney (arrows). B CDFI showed a lack of intratumoral vessel signal in the mass (arrows). C CEUS imaging in the initial enhancement: the tumor enhanced simultaneously with the cortex (arrows). D CEUS imaging at the peak enhancement: the tumor showed homogenous isoenhancement similar to the peritumoral cortex (arrows). E CEUS imaging at the medullary phase showed prolonged enhancement (arrows). F Unenhanced CT demonstrated a slightly hypodense (-14 HU) mass in comparison with renal parenchyma, with scattered visible fat (arrow). G Corticomedullary phase: mildly heterogeneous enhancement (65HU) was noted (arrow). H Nephrographic phase: the tumor’s mean density value gradually increased to 79 HU (arrow). I Tumor’s mean density value decreased to 43 HU in the delayed phase (arrow)

Full size image
Fig. 2
figure 2

A 49-year-old woman with a 2.8-cm diameter atypical AML was misdiagnosed as RCC using both CEUS and CECT. A Conventional ultrasound demonstrated an exophytic hypoechoic mass located in the upper pole of the left kidney (arrows). B CDFI showed a lack of intratumoral vessel signal in the mass (arrows). C CEUS imaging in the initial enhancement: the tumor enhanced simultaneously with the cortex (arrows). D CEUS imaging at the peak enhancement: the tumor showed mildly heterogeneous isoenhancement (arrows). E CEUS imaging at the medullary phase: the tumor showed faster wash out than the renal cortex (arrows). F Unenhanced CT demonstrated a hyperattenuating (38 HU) mass in comparison with renal parenchyma, without visible fat (arrow). G Corticomedullary phase: the mass showed heterogeneous enhancement (102 HU) (arrow). H Nephrographic phase: the tumor’s mean density value gradually decreased to 98 HU (arrow). I Tumor’s mean density value decreased to 69 HU in the delayed phase (arrow)

Full size image
Fig. 3
figure 3

A 60-year-old man with a 3.2-cm diameter ccRCC was diagnosed using both CEUS and CECT. A Conventional ultrasound demonstrated a heterogeneous mass with tiny intratumoral cysts located in the lower pole of the right kidney (arrows). B CDFI showed some blood flow signals around the tumor periphery and a few strip-like blood flow signals within the tumor (arrows). C CEUS imaging in the initial enhancement: the tumor enhanced simultaneously with the cortex (arrows). D CEUS imaging at the peak enhancement: the tumor showed heterogeneous isoenhancement with many unenhanced areas (arrows). E CEUS imaging showed that the tumor wash out in the medullary phase was faster than that of the renal cortex (arrows). F Unenhanced CT demonstrated a mildly heterogeneous mass (23 HU) (arrow). G Corticomedullary phase: the mass showed heterogeneous hypoenhancement (76 HU) (arrow). H Nephrographic phase: the tumor mean’s density value increased to 108 HU (arrow). I Tumor’s mean density value decreased to 75 HU in the delayed phase (arrow)

Full size image
Fig. 4
figure 4

A 63-year-old man with a 1.9-cm diameter pRCC was diagnosed using CEUS but was misdiagnosed using CECT as AML without visible fat. A Conventional ultrasound demonstrated a mildly hyperechoic mass located in the interpolar pole of the right kidney (arrows). B CDFI showed a lack of intratumoral vessel signals in the mass (arrows). C CEUS imaging in the initial enhancement: the tumor enhanced slower than the renal cortex (arrows). D CEUS imaging at the peak enhancement: the tumor showed homogenous hypoenhancement (arrows). E CEUS imaging at the medullary phase: the tumor showed faster wash out than the renal cortex (arrows). F Unenhanced CT demonstrated a homogenous mass (14 HU) (big arrow). G Corticomedullary phase: the mass showed mildly heterogeneous hypoenhancement (16 HU) (big arrow). H Nephrographic phase: the tumor’s mean density values was 41 HU (big arrow). I Tumor’s mean density values decreased to 37 HU in the delayed phase (big arrow). The other lesions were simple cysts (small arrows)

Full size image
Table 4 Comparison of qualitative diagnostic accuracy on different histotypes of small renal masses between CEUS and CECT
Full size table

Diagnostic ability of CEUS and CECT

The sensitivity, specificity, PPV, NPV, and AUC with 95% CI for the diagnostic ability of CEUS and CECT are summarized in Table 5. Both CEUS and CECT have good diagnostic performance in the differential diagnosis of benign and malignant SRMs with sensitivity of 93.5% and 89.2%, specificity of 68% and 76%, PPV of 91.6% and 93.3%, NPV of 73.9% and 65.5%, and AUC of 0.808 and 0.826, respectively. Both these methods demonstrated good diagnostic performance in the differential diagnosis of benign and malignant SRMs, and there was no statistically significant difference in any of the diagnostic performance indices between these two methods (P > 0.05)

Table 5 Diagnostic ability of CEUS and CECT on benign and malignant small renal masses
Full size table

Discussion

The most common benign renal tumors include AML and oncocytoma. RCC is the most common malignancy, accounting for 90–95% of kidney cancers with three major subtypes: clear cell (70%), papillary (10–15%), and chromophobe (5%). Gender distribution shows a higher incidence in females for AML and males for RCC, which was also reflected in our results. CCRCC is present at a higher rate of metastasis mostly with hypervascular feature, whereas pRCC and chRCC are considered less aggressive renal tumors and are typically hypovascular [22]. While standard treatment for RCC used to be radical nephrectomy, clinically localized RCC, when amenable, is now treated by nephron-sparing surgery, with similar oncologic outcomes and better functional outcomes [23]. In our study, benign lesions and low-aggressive RCCs were mostly managed with nephron-sparing partial nephrectomy. Many studies have reported that both CEUS and CECT imaging technologies play important roles in diagnosing renal tumors; however, to our best knowledge, this is the first study to investigate the comparison between CEUS and CECT in differentiating benign from malignant SRMs.

On CEUS and CECT imaging, there are some distinct qualitative features characterizing RCC, such as peak tumor enhancement, tumor heterogeneity, fast wash out in the late phase, and perilesional rim-like enhancement. While heterogeneous hyperenhancement, wash out in the late phases, and the presence of a pseudocapsule often predicted ccRCCs, AMLs more often showed homogeneous iso- or hypoenhancement and prolonged enhancement [19, 24–27]. Xu et al. [25] found that using early wash-out enhancement and heterogeneous enhancement or peritumoral rim enhancement as the diagnostic standard of CEUS for RCC, the corresponding sensitivity, specificity, PPV, NPV, and accuracy were 88.2%, 97.0%, 98.8%, 74.4%, and 90.5%, respectively, and regardless of the subtype of RCC found, inhomogeneous enhancement and pseudocapsule enhancement were the major findings. However, for tumors smaller than 3 cm, homogeneous enhancement was more frequently seen on both CEUS and CECT, regardless of the histologic subtype [14, 27], because the small tumors grow slowly and rarely have necrosis change. For the continually growing tumor with a lack of blood supply, necrotic areas will be observed frequently. In our study, we also found that most benign and malignant SRMs demonstrated homogeneous enhancement on both CEUS and CECT, limiting the utility of tumor heterogeneity in differentiating benign and malignant SRMs. Other features such as peritumoral rim enhancement, hypervascularity, and fast wash out were considered the most commonly observed enhancement characteristics for small-size RCC [14, 28]. Although these two methods have proven useful for distinguishing benign from malignant renal tumors, because of the multiple histotypes of renal masses, the overlapping imaging features and density values in different histological types render their accuracy in characterization of SRMs still controversial.

The main difficulty of imaging renal masses is to differentiate malignant lesions, most often RCC, from benign entities such as AML and oncocytoma. Most ccRCCs are hyperenhancing and show peak enhancement earlier than renal parenchyma, but the imaging features can overlap with oncocytomas, and there can also be overlap between hypovascular benign and malignant renal lesions, such as papillary RCCs and lipid poor AMLs [16]. In the 75 ccRCCs in our study, 66 were demonstrated by both CEUS and CECT, 5 were misdiagnosed by CEUS as AMLs, three were misdiagnosed by CECT as AMLs, and 1 was misdiagnosed by both CEUS and CECT as an AML. Oncocytomas are considered benign tumors that have low metastatic potential but their growth rate is similar to RCC [29]. Imaging diagnosis of oncocytomas can be difficult, and some studies have concluded that these lesions cannot be efficiently differentiated by CEUS and CECT [30]; thus, oncocytoma remains the most commonly excised benign solid renal mass. In the present study, the three oncocytomas were all misdiagnosed by both CEUS and CECT as RCCs. The classic AMLs could be easily demonstrated on conventional CT because of their macroscopic fat, a specific finding rarely seen in other solid renal masses. However, approximately 5% of AMLs have insufficient fat to be detected and are generally considered indistinguishable from RCC or oncocytomas at imaging due to their overlapping imaging characteristics [31, 32]. In the 20 AMLs of our study, 15 (10 classic AMLs and 5 lipid poor AMLs) were correctly diagnosed by CEUS and CECT, while the other 5 lipid poor AMLs were misdiagnosed as RCCs by either or both methods. In our study, we found that there was no significant difference in the qualitative diagnostic accuracy of small AMLs and ccRCC between using CEUS and CECT.

Slow wash-in and fast wash-out patterns and hypoenhancement at peak are considered the more common features that could be helpful in the differentiation of pRCC with ccRCC on CEUS [13, 21, 33]. Similarly, on CECT, pRCC has reportedly demonstrated a significantly lower degree of enhancement in the corticomedullary phase [34]. However, for small-size tumors, it is difficult to determine the true enhancement of the lesion using CECT because of its limitations: partial volume effect and pseudoenhancement [35]. CEUS imaging technology is more sensitive and accurate than CECT in detecting microvascularity of tumors because the ultrasound contrast agent is purely intravascular [36]. In this study, five pRCC lesions demonstrated equivocal or no enhancement by CECT, whereas the tumor enhancement in all of these lesions was detected by CEUS. The slow wash-in and fast wash-out patterns were detected more easily using CEUS than CECT, because CECT provides intermittent static acquisitions, while CEUS provides a complete and continuous dynamic progress of contrast agent perfusion and clearance information in tumors. A previous study suggested that CEUS may be helpful for diagnosing hypovascular renal lesions that demonstrate equivocal or no enhancement by CECT [13]. In this study, we found that the qualitative diagnosis of pRCC using CEUS was significantly better than using CECT.

In our study, we used the diagnostic criteria of CEUS and CECT according to the established criteria developed on the basis of enhancement patterns of renal masses previously described. Excellent inter-reader agreement was found for the differential diagnosis of benign and malignant SRMs by both CEUS and CECT. Both these two methods demonstrated good diagnostic performance with sensitivity of 93.5% and 89.2%, specificity of 68% and 76%, PPV of 91.6% and 93.3%, NPV of 73.9% and 65.5%, and AUC of 0.808 and 0.826, respectively, although there was no statistically significant difference in any of the diagnostic performance indices between these methods.

This study had some limitations. First, this was a retrospective study. Second, only the masses confirmed by histopathologic diagnosis after surgical resection were enrolled in this study, which might have some degree of selection bias. Many benign lesions at follow-up without histopathologic diagnosis were excluded; thus, the sample of AMLs was relatively smaller than that of RCCs, the follow-up studies maybe enrolled in further study. Finally, this study included few oncocytomas, pRCCs, and chRCCs. Because of the small number of renal masses, this study did not evaluate differences in each histotype of renal masses, and a larger sample size is needed for further investigation.

In conclusion, our study suggests that both CEUS and CECT imaging modalities are effective in the differential diagnosis of benign and malignant SRMs. Furthermore, CEUS may be more effective than CECT in the qualitative diagnosis of small papillary RCC.