Value of elastography in characterization of solid renal masses

Abstract

Background

Solid renal masses often come to light as incidental findings during abdominal ultrasound examinations. Once detected, determining whether a renal mass is benign or malignant becomes imperative for informed decision-making regarding management and treatment strategies. In this investigation, the aim was to explore the diagnostic efficacy of real-time strain sonoelastography in assessing solid renal masses.

Methods

This prospective research was steered on 26 individuals diagnosed with a solid renal mass, as endorsed by pathological analysis after surgical removal or biopsy. Elastography was performed on all patients. The measurement of strain index values for tissues was achieved by placing regions of interest of equal or near-equal size on both the tumor (A) and the adjacent normal renal cortex (B).

Results

Strain elastography showed no correlation with patient’s age, size of mass and probe to mass distance with p > 0.05 all. Sensitivity analysis showed that strain index can significantly predict malignant renal masses (P = 0.003) using a cut-off point 2, with 92.9% area under curve, 95.2% sensitivity, 80% specificity, 80% negative predictive value, 95.2% positive predictive value, and overall diagnostic accuracy 92.3%. Strain index > 2 was an independent predictor for malignant renal masses (P = 0.025), odds ratio 7.29 when adjusting for other risk factors. Malignant renal masses were significantly higher strain index compared to benign lesions with (P = 0.001).

Conclusions

Strain elastography is a valuable technique for distinguishing between malignant and benign solid renal tumors. Benign lesions have lower strain index values compared to malignant ones, making the strain index a useful screening tool for distinguishing between benign and malignant renal masses using cut-off point 2.

1 Introduction

Ultrasound scanning of the abdomen often leads to the unintentional detection of solid renal tumors [1]. In order to identify a renal mass, it is necessary to classify the tumor as either benign or malignant for concluding appropriate strategies for its care and treatment. Non-cancerous renal masses can be controlled with monitoring using imaging methods or may require biopsy, however cancerous renal masses necessitate surgical excision. The most commonly observed malignant tumor in the kidney is renal cell carcinoma. It accounts for around 2% of all malignancies in adults [2].

Renal oncocytoma and angiomyolipoma are the two most commonly observed benign tumors that impact the kidney. Ultrasonography is commonly employed as an accidental diagnostic tool for renal malignancies [3]. Next, the procedure involves the exploitation of computed tomography (CT) and/or magnetic resonance imaging (MRI) with a contrast agent. Nevertheless, distinguishing between these masses can provide challenges due to the resemblances in their imaging characteristics. The literature has identified various imaging variations of both non-cancerous and cancerous kidney masses. However, it provided evidence indicating that 6.9% of individuals in whom renal cell carcinomas were suspected and surgical nephrectomy was performed were found to have pathologically confirmed angiomyolipomas [4]. Thus, preoperative differentiation is important as the management differs between these tumors. Additionally, CT and MRI have certain drawbacks. Some of the potential risks associated with CT examination include high doses of radiation, claustrophobia in MRI, and kidney effects of contrast agents in both procedures. Despite the clear diagnostic advantages associated with percutaneous biopsy, certain problems may arise, including bleeding and the development of arteriovenous fistula, which have the potential to be life-threatening. Additionally, tumor seeding represents another drawback associated with biopsy. Therefore, there is a need for an additional non-invasive approach to differentiate between non-cancerous and cancerous lesions [5].

A novel imaging technology known as ultrasound elastography has been developed to quantify the biological tissues' pliability or rigidity. Strain elastography is a procedure that measures the hardness of a mass within an organ by comparing it to the surrounding normal organ tissue. The strain index value obtained is a measure of the balance between the hardness of the mass and the normal parenchyma. A variety of benign and malignant tumors in the thyroid gland, breast, prostate, testes, and liver have been identified using the strain elastography approach [6, 7].

A novel method called strain elastography may measure the elasticity of tissues by means of controlled compressions and decompressions. There are a number of studies that discuss the effectiveness of elastography for solid lesions of the kidneys, placenta, and breasts [5, 7, 8].

In this investigation, the aim was to explore the efficacy of real-time strain sonoelastography in the diagnosis of solid renal masses.

2 Patient and methods

This prospective research was conducted on 26 individuals diagnosed with a solid renal mass, as endorsed by pathological analysis after surgical removal or biopsy. The research was done from October 2021 and June 2022 after approval from the Ethical Committee Cairo University Hospitals, Cairo, Egypt (approval code: MS-511-2021). An informed written consent was obtained from the cases.

Three individuals were omitted from the study because of cystic components found in their renal masses, while three more were omitted due to significant alterations in backpressure on the kidneys.

In the supine and lateral decubitus positions, patients were examined using a Canon Aplio a550 Ultrasound machine that was programmed to do strain elastography as well as B-mode sonography. A convex 3.5–5 MHz multifrequency transducer was used for transabdominal imaging. Elastography was done by a blinded single operator, one day prior to surgical resection of the tumor.

Prior to conducting an elastographic examination, the morphologic characteristics of the lesions, including their size, shape, and contour, were assessed in grayscale. The patients were positioned on their sides, across from the kidney mass. After taking a deep breath and holding it, the examiner applied vertical pressure to the patient's back using the ultrasound probe. This was done while the examiner could see the kidney mass and its surrounding cortex in the ultrasound image. In response to compression and decompression, the mass and kidney were found to move in real time. The imaging box was adjusted during strain elastography so that it completely covered the mass and the surrounding renal cortex (Figs. 2,3). As the probe moved over the masses, grayscale sonograms appeared on the screen next to the elastographic pictures.

Elastography classified tissues based on strain, which was defined as the extent to which tissues were displaced in reaction to applied pressure. When subjected to compression and decompression, tissues under high strain were more displaced than those under low strain. Red denotes the most severely stressed tissues, green those with an intermediate strain, and blue those with the least stressed tissues (hard tissues), in that order [9] (Figs. 2,3).

After 10 or 12 cycles of compression and decompression, the ultrasound machine automatically generated elastographic images by comparing two neighboring frames taken when the tissue was compressed and decompressed. Elastographic waveforms depicting compression and decompression, with sinusoidal forms above and below the waveform scale baseline, respectively, were displayed on the screen [10] (Figs. 2,3).

Since there was no external pressure while the measurements were taken, they solely capture the internal dynamics of the decompression period [11].

2.1 Strain index

Placing regions of interest (ROIs) of identical or nearly equivalent size on the tumor (A) and nearby normal renal cortex (B) allowed us to evaluate the strain index values of tissues. When feasible, non-necrotic portions of the masses should have been used to implant the tumor and renal cortex ROIs at the same depth [7]. Automatically, the sonography equipment computed the strain index value (B/A), which reflects the lesion's stiffness, by comparing the tumor (A) to the nearby normal renal cortex (B). If the lesion is more rigid than the surrounding cortex, the strain index will be higher [5]. Once the strain index values were measured, researchers compared the average values for benign and malignant tumors. To further investigate the potential impact of patient age, tumor size, and probe-tumor distance (the distance between the renal tumor's surface and the skin surface) on strain index value readings, we also compared benign and malignant tumors. A PACS system was subsequently utilized to acquire and store the images for later viewing.

2.2 Statistical analysis

Analyses were carried out using SPSS 22nd edition. For parametric variables, quantitative data were given as mean and standard deviation. For non-parametric variables, median and range were used. Student’s T test and Mann–Whitney U test were used for comparisons, respectively. We used a Chi-square test to compare the two groups' categorical data, which were given as a percentage or frequency table. The diagnostic capacity of the strain index for kidney tumors was evaluated by a sensitivity analysis. Predictor variables for kidney tumors were evaluated using a binary logistic regression model. To determine the relationship between strain index and other numerical variables, the Spearman correlation test was employed. Significant was any p value less than 0.05.

3 Results

A mean age of 48 ± 14.6 years was observed. Out of the total, 14 individuals (53.8%) were male and 12 (46.2%) were females. In the present study, renal masses had a median size in the maximal diameter 5.8(3.5–10) cm, median probe to mass distance 3.3(0.9–9.5) cm and median strain index 3.8(1.02–5.6). Regarding histological diagnosis, 21(80.8%) renal masses were malignant while only 5(19.2%) were benign in nature, including angiomyolipoma in two cases, oncocytoma in 3 cases (Table 1).

Table 1 Demographics, ultrasound strain elastography findings and histopathological diagnosis among the included patients

Malignant renal masses had significantly higher strain index compared to benign lesions (p = 0.001), while insignificantly different was reported between study groups in terms of age, gender, size, and probe to mass distance (p > 0.05) (Table 2).

Table 2 Comparison of demographics and ultrasound findings according to diagnosis

The results of the sensitivity assessment indicated that strain index can significantly predict malignant renal masses (p = 0.003) using a cut-off point 2, with 92.9% AUC, 95.2% sensitivity, 80% specificity, 80% negative predictive value, 95.2% positive predictive value, and overall diagnostic accuracy 92.3% (Fig. 1).

Fig. 1

ROC curve showing predictability of strain index for malignant renal masses

Strain elastography showed no correlation with patient’s age, size of mass and probe to mass distance with p > 0.05 all (Table 3).

Table 3 Correlation matrix between strain elastography and other confounding factors

Strain index > 2 was an independent predictor for malignant renal masses with p value 0.025, odds ratio 7.29 when adjusting for other risk factors (Table 4).

Table 4 Binary regression model showing predictors for malignant renal masse

Case 1: A 56-year-old male patient appeared with a tumor in the right kidney. CT showed right renal partly exophytic lesion measuring 5.4 × 4.2 cm and showing faint post-contrast enhancement. Strain index: 3.23. Histopathology showed papillary renal cell carcinoma (Fig. 2).

Fig. 2

Strain elastography for papillary renal cell carcinoma

Case 2: A 47-year-old male patient exhibited a tumor in the left kidney. CT showed left renal mid and lower zones fairly defined mass lesion measuring 11 × 5 cm with heterogenous post-contrast enhancement and central hypodense area of breaking down. Strain index: 2.13. Histopathology showed clear cell renal cell carcinoma (Fig. 3).

Fig. 3

Strain elastography for clear cell renal cell carcinoma

4 Discussion

The renal mass biopsy is widely regarded as the most reliable method for diagnosing benign or indolent disease before surgery. However, its use has been hindered by a non-diagnostic rate of 10–15% and worries about variations within the tumor [12, 13].

Regarding renal masses imaging, contrast-enhanced computed tomography (CECT) is now regarded as the preferred method for evaluating solid kidney tumors. Malignancy is linked to the presence of post-contrast enhancement and heterogeneity in a renal lesion as well as to some extent, can be utilized to differentiate between RCC subtypes, but with limited efficacy [14, 15].

While not obligatory, the corticomedullary and excretory stages may aid in the classification of RCC, distinguishing it from urothelial carcinoma, or assessing the presence of the tumor in the collecting system [16]. The thickness of the slices is a crucial factor in a renal mass CT approach. It is advisable to use thin slices, measuring 2–3 mm, in order to minimize partial volume artifacts that might impact the characterization of tiny renal masses [17].

T2-weighted images helps to find and describe cystic kidney masses, should be included in a conventional renal mass MRI procedure. The T2 signal strength of a solid renal mass can also be used to rule out certain histology [18].

It is advisable to employ T1-weighted gradient-echo in-phase and opposed-phase imaging techniques for the detection of macroscopic or microscopic/intracytoplasmic fat, as well as hemosiderin, within a renal mass. It is advisable to do T1-weighted three-dimensional fat suppressed gradient-echo imaging before and after the delivery of gadolinium-based contrast material during the corticomedullary, nephrographic, and excretory phases. The enhancing pattern of a renal tumor on multiphase MRI has been shown to be advantageous in differentiating between various subtypes of renal RCC [19].

The use of diffusion-weighted MRI is a prevalent elective modality within renal mass therapy. The use of apparent diffusion coefficient (ADC) measurements has been proposed in many studies as a means to enhance the characterization of renal masses [20].

CT and MRI have several drawbacks. Some of the potential risks associated with CT examination include high doses of radiation, claustrophobia in MRI, and kidney effects of contrast agents in both procedures. Despite the clear diagnostic advantages associated with percutaneous biopsy, some problems may arise, including bleeding and the development of arteriovenous fistula, which have the potential to be life-threatening. Additionally, tumor seeding is another drawback associated with biopsy. Therefore, there is a need for further non-invasive techniques to distinguish between benign and malignant tumors [5].

One symptom of diffuse parenchymal diseases is a change in tissue stiffness, which ultrasound elastography can identify, as well as alterations in tissue architecture [7].

In the current study we found that malignant renal masses showed a significantly higher strain index with median SI 3.9 (range 1.83–5.6) compared to benign lesions SI 1.36 (range 1.02–3.8), sensitivity analysis showed that strain index can significantly predict malignant renal masses using a cut-off point 2, with AUC 92.9%m 95.2% sensitivity, 80% specificity, 80% negative predictive value, 95.2% positive predictive value, and overall diagnostic accuracy 92.3%. These findings were consistent with that of Onur et al. [7] who found that the mean strain index value for kidney masses that were malignant (4.05 ± 2.17) was significantly higher than the value for benign masses (1.43 ± 0.94; P < 0.05). A separate investigation carried out by Tan et al. [21] stated that strain ratio values evaluated by radiologist 1 and radiologist 2 in angiomyolipomas were 0.15 ± 0.06 and 0.18 ± 0.09, respectively, and 0.64 ± 0.15 and 0.63 ± 0.19 in renal cell carcinomas. The elasticity patterns and strain ratios of benign renal masses and renal cell carcinomas differed significantly.

Another cross-sectional study exposed that there was a statistically significant increase in strain ratio values in the malignant masses compared with the benign masses found also that sensitivity analysis demonstrated that a strain ratio greater than 2.295 was identified as the most effective threshold value. The diagnostic accuracy for identifying a malignant mass was found to be 93.8% for sensitivity, 80.5% for specificity, 65.2% for positive predictive value, and 97.1% for negative predictive value [22]. Other studies had settled a lesser cut-off point for strain ratio as Fukuda et al. [23] showed that sensitivity analysis revealed that the most effective threshold for distinguishing between RCCs and benign masses is 1.29, with a sensitivity of 85% and specificity of 100%. A value of 0.88 was observed for the area under the curve, which is attributed to the inclusion of individuals with minor renal masses. Additionally, Tezcan et al. [5] reported in their cross-sectional study that average strain index for malignant lesions was found to be substantially greater than that of benign lesions, with a cut-off value of 0.3. The sensitivity and specificity were determined to be 100% and 68.7%, respectively. The positive and negative predictive value for identifying malignant lesions were found to be 80% and 100%, respectively, with an area under the curve (AUC) of 0.899.

To our knowledge this is the first study to conduct logistic regression model to evaluate predictors for malignant renal masses when we found that strain index > 2 was an independent predictor for malignant renal masses (P = 0.025), odds ratio 7.29 (95% CI 1.3–41.4) when adjusting for other risk factors.

Limitations of this study including that small sample size and relatively low incidence of renal malignancies in Egypt which made the variety of diagnosis decreases. Some of the lesions have no histopathologic diagnosis and had to be excluded from the study. The nature of cross-sectional study takes only information about the current situation without adequate assessment of predictors and confounding factors and lack of randomization. Furthermore, ultrasonography, being a tool that relies on the operator, and strain elastography, which is highly reliant on the operator's skill level, raised concerns about the repeatability of results. However, we successfully addressed this issue by employing a single operator for all instances.

5 Conclusions

Strain elastography is a valuable technique for distinguishing between malignant and benign solid renal tumors. Benign lesions have lower strain index values compared to malignant lesions, hence enabling the utilization of strain index as a screening technique for distinguishing between benign and malignant renal masses by employing a cut-off point of 2. Given the ease of application, non-invasive nature, and cost-effectiveness of strain elastography as an imaging tool, we propose that its integration with traditional sonography might enhance accuracy and aid in the differential diagnosis of these cancers.