Role of non-contrast magnetic resonance imaging in pre-surgical evaluation of renal masses in renal impairment patients

The aim of this work is to study the role of non-contrast MRI in pre-surgical evaluation of renal masses in renal impairment patients as confirmed by both intraoperative and histopathological findings. Intraoperative and histopathological findings were correlated with radiological data. This prospective study included 20 patients comprising 25 renal masses. The data were collected in a period from April 2018 to September 2019. All patients underwent partial or radical nephrectomy by the same surgeon. Based on MRI findings, 9 masses (36%) and 8 masses (32%) were found to be associated with collecting system invasion and perinephric fat invasion, respectively. Histopathological assessment confirmed only 6 cases (24%) with collecting system invasion and 7 cases (28%) demonstrated perinephric fat. Seven masses (28%) had intralesional hemorrhage detected by MRI and confirmed by pathological findings. The MRI detected 6 cases (24%) with lymph nodes invasion, while the histopathological assessment confirmed lymphatic invasion in 7 cases (28%). Only 2 cases (8%) had vascular invasion detected by preoperative MRI and confirmed by histopathology and surgery. The final histopathological examination confirmed 20 malignant neoplasms (80%: RCC = 18, leiomyosarcoma = 2), 3 benign neoplasms (12%: angiomyolipoma = 1, oncocytoma = 2) and 2 non-neoplastic benign masses (8%: renal abscess = 1, xanthogranulomatous pyelonephritis = 1). Non-contrast MRI is a crucial imaging tool in renal impairment patients who cannot be examined with contrast-enhanced CT or MRI. It assesses the extent of the renal sinus fat and the perinephric fat invasion.


Background
Recently, as a result of increased use of high-resolution cross-sectional imaging, the number of incidentally discovered solid renal masses has increased, requiring more imaging characterization for proper diagnosis [1]. This characterization extends beyond the simple determination of malignancy and benignity, as it determines and guides the therapeutic approach and follow-up management [2].
In addition to enabling proper characterization of solid renal lesions, cross-sectional imaging is also necessary for staging purposes in patients with known disease, for assessment of recurrence in patients with previously treated disease, and for active surveillance in high-risk patients [3].
RCC is one of the most frequently diagnosed adult cancers, accounting for 2-3% of all adult malignancies [4]. There are two main benign lesions that may be difficult to differentiate from RCC. These are angiomyolipoma (AMLs), in particular the lipid-poor subtype, and oncocytoma, which represent 3-7% of all renal tumors [5].
Because of the heterogeneity of imaging features as well as overlapping imaging characteristics of renal masses, Zaytoun et al. Afr J Urol (2021) 27:61 identifying reliable imaging criteria for recognition of malignant versus benign masses remains a challenging task. Several imaging parameters have been proposed to distinguish the different renal lesions. Abdominal ultrasonography and computed tomography (CT) are commonly used for a variety of renal indications. MR imaging can be particularly helpful when renal lesions are detected [6].
Traditionally, patients with impaired renal function could not be imaged with a CT scan with iodinated contrast media (ICM) due to the risk of contrastinduced nephropathy (CIN). These patients were alternatively imaged by MRI with gadolinium. However, the recent identification of the association between nephrogenic systemic fibrosis (NSF) and gadolinium administration has created significant challenges for radiologists when faced with the need for evaluation with contrast-enhanced cross-sectional imaging. As a result, the use of non-contrast MRI is thought to be the solution for renal impairment patients [4]. The aim of this work is to study the role of non-contrast MRI in pre-surgical evaluation of renal masses in renal impairment patients.

Methods
The study included 20 patients with renal impairment presenting with renal mass detected by ultrasonography or non-contrast CT and referred our institute for noncontrast MRI evaluation before surgical intervention (partial or radical nephrectomy) performed in the urology department.
The medical ethics were considered including informed consent that was obtained from all cases included in the study after explaining the purpose of the study. Confidentiality was maintained throughout the study.
Initial routine evaluation included thorough history, clinical examination and serum creatinine and eGFR assessment. Operative data were obtained and correlated with the radiological data. Histopathological analysis for the surgical specimens obtained for all patients and correlated with the radiological data and considered as the reference gold standard.

MR Imaging protocol
MRI sequences were acquired on Ingenia 3Tesla closed magnet MRI machine (Philips Medical systems, Netherlands), as well as Ingenia 1.5Tesla closed magnet MRI machine (Philips Medical systems, Eindhoven, The Netherlands). The choice of either machine was entirely based on availability, and comparison between the results of each was beyond the scope of this study. Axial T1-weighted, axial and coronal T2-weighted images as well as axial diffusion-weighted images were acquired, and apparent diffusion coefficient (ADC) maps were reconstructed on the work station.

Results
The study included 20 patients with a total number of 25 renal masses (3 patients had bilateral renal masses and 2 patients had two masses in the same kidney). The surgical intervention encompassed 13 radical nephrectomy and 10 partial nephrectomy. All surgeries were performed by the same surgeon in the urology department. Table 1 shows the final histopathological diagnosis of the renal masses included in the study.

Non-contrast magnetic resonance imaging (MRI) findings and correlation with surgical and pathological findings
1 Distribution of the renal mass Figure 1 summarizes the distribution and location of renal masses as evaluated by preoperative non-contrast MRI. 2 Composition of the mass and presence of intralesional hemorrhage and necrosis Eleven lesions were seen entirely solid (44%), 3 cystic lesions (12%), and eleven solid lesions with necrotic areas (44%) (Fig. 2) ( Table 2).
Among the 25 masses, 7 (28%) masses showed intralesional hemorrhage detected by MRI and confirmed by pathological findings (Table 3) (Fig. 3a). Masses with hemorrhagic content are detected in the MRI as areas of hyperintensity in T1-weighted images.

Assessment of invasion of renal sinus fat and collecting system
Out of 8 masses found to have renal sinus fat invasion on preoperative MRI (Fig. 2), surgical findings confirmed only 6 of them (24%) and the pathological findings stated that only 5 cases (20%) had renal sinus fat invasion (Table 4). Preoperative MRI showed that 9 masses had collecting system invasion: out of those only 7 and 6 cases were confirmed by surgical and pathological finding, respectively (Table 4). 4 Assessment of invasion of renal capsule and perinephric space Surgical and pathological data confirmed the presence of renal capsule and perinephric fat invasion in 7 cases out of 8 cases was initially spotted on MRI ( Fig. 2). (Table 4).

Invasion of the lymph-vascular space
Six cases were found to have enlarged para-aortic lymph nodes on MRI (Fig. 2 b, c). However, the surgical and the pathological findings confirmed seven cases to be associated with enlarged para-aortic lymph nodes (2: reactive lymphoid hyperplasia, 5: lymphatic metastasis) ( Table 4). Imaging, surgical findings and pathology confirmed two cases associated with vascular invasion: one of them had tumor thrombus in the ipsilateral renal vein and the other had renal vein invasion and renal artery encasement with clear IVC (Table 4).

Diffusion restriction
Eighteen masses (72%) were found to be restricted in diffusion-weighted sequences (DWI) (Fig. 3 b). ADC values were measured in the restricted areas (Fig. 3  c). The region of interest (ROI) was fitted to area only intended to be measured to avoid averaging from surroundings. Its size averages 0.5 cm 2 ; however, in larger lesions; larger ROI size was used. ROI was fitted in a homogenous area and was measured in multiple sites in the restricted area, and the lowest mean value was taken. ADC values in the 18 restricted masses range from 0.5-1.4 × 10 -3 mm 2 /s (diffusion restricted masses), and the mean ADC value was 0.88 × 10 -3 mm 2 /s. Seven masses (28%) showed free diffusion with higher ADC values more than 2 × 10 −3 mm 2 /s ranging from 2.0-2.5 × 10 -3 mm 2 /s (diffusion-free masses).

Discussion
Non-contrast MRI is a useful noninvasive imaging tool for diagnosis, characterization and differentiation of the most common types of renal masses in patients with renal impairment. It provides excellent soft tissue contrast and functional information. Renal MR imaging allows characterization of lesion location, extent, lymphovascular invasion and diffusion restriction and may allow differentiation of malignant from benign lesions. In addition, non-contrast MR imaging may aid in classifying RCC subtypes. In our study, RCC was the most prevalent tumor subtype accounting for 90% of cases. This is similar to the reported data in the literature [4]. Regarding RCC subclassification, Ramamurthy et al. and Chivelle et al. stated that clear cell RCC is the most common subtype of RCC [7,8]. Papillary carcinomas are the second most common subtype of RCC. They are bilateral in 4% of cases [8]. Chromophobe RRC is the third most common subtype [8]. This categorization coincides with the findings in our study.
Muglia et al., [9] reported that on T1-weighted images, clear cell RCC shows low to similar signal intensity when compared to that of the renal parenchyma. This comes in concordance with our results. Clear cell RCC seen in our patients showed iso-to hyperintense signal on T2w images, with hypo-to iso-intense signal on T1w images, and six out of eight clear cell RCC masses (75%) were heterogeneous showing central break down, which matches the literature stating that clear cell RCC tends to be heterogeneous with areas of necrosis [9,10].
In our study, clear cell RCC masses with central break down showed central irregular area of T2-weighted   Necrosis has been shown to correlate with tumor size and grade [10]. Some authors demonstrated that the smaller papillary RCCs are homogenous solid tumors, but when the tumor exceeds 4 cm in diameter, internal heterogeneity may be observed because of hemorrhage, calcification, and necrosis [9,11]. This was not the case in all of the papillary RCC tumors included in our study, as we found one papillary RCC mass measured 7.7 cm in the largest dimension, still showed homogenous texture with no internal hemorrhage or necrotic changes. The other small papillary RCC masses measured less than 4 cm were all still homogenous lesions.
Vendrami et al. [12] stated that MR imaging features of papillary RCC in most cases show well-defined circumscribed homogenous peripherally located small sized (3 cm or less) tumors, with a surrounding fibrous capsule showing low signal intensity in both T1 and T2 weighted images. This is not matching with the finding in our study, as we encountered 5 papillary RCC masses, 3 of them showed the features stated by the literature, and the other 2 masses: one large sized (7.7 cm) homogenous mass with no internal hemorrhagic component and the other one was large sized mass (7 cm) with hemorrhagic content. No hypointense capsule was noted around any of the previously mentioned papillary RCC masses included in the study.
Intratumoral hemorrhage may occur and has a variable appearance depending on the stage of degradation of the blood products. Subacute to chronic hemorrhage generally demonstrates high signal intensity on both T1-and T2-weighted images, stated by Vendrami et al. [12]. This is comparable to the findings in our study, where renal masses with intralesional hemorrhagic contents showed T1 hyperintense signal with intermediate or low signal intensity on T2w images, so in our study, T1w sequence was the most important sequence for identifying intralesional hemorrhagic content [12].
Chromophobe RCC tends to be well-circumscribed and homogenous with no cystic changes or central necrosis even in large tumors; they show low to intermediate T2 signal intensity [9,10]. This matches the MRI findings of the 2 cases of chromophobe RCC in our study; they were solid homogenous masses with intermediate T2 signal intensity.
There is consistent association between some MR imaging features and specific histological subtypes of RCC. Similarly, we found greater prevalence of high signal intensity on T2W images in patients with clear cell RCC, contrasting with lower signal intensity found in those with papillary RCC, chromophobe RCC [13][14][15][16].
Wang et al. [17] stated that RCC typically shows restricted diffusion, which is high signal intensity in DWI with corresponding low signal in ADC maps. These findings match the results obtained from the current study where all cases of RCC showed restricted diffusion. DWI can be helpful in detecting renal masses, especially in patients who cannot receive gadolinium contrast material. The presence of restricted diffusion and lower ADC values can be seen in masses that are both malignant and benign, including RCC, oncocytoma, AML, and renal abscess [12]. This matches with what we found in our results, as among the 20 malignant lesions encountered in our study, 16 (80%) masses showed areas of restricted diffusion and 4 (20%) masses demonstrated free diffusion (3 of them were papillary RCC and 1 clear cell RCC), and the benign tumors all showed free diffusion, while the 2 infective masses were restricted (XGPN and renal abscess). Based on these results, the presence of diffusion restriction cannot be considered a reliable feature for the malignant renal masses.
Regarding AML, our study showed that the gross macroscopic fat content appeared as T1 hyperintensity which was suppressed in the fat suppression sequence. Similarly Vendrami et al. [12] stated that classic AMLs (lipid-rich AML) demonstrate high T1 signal intensity because of the presence of macroscopic fat content. The presence of macroscopic fat can be appreciated as suppression of signal intensity on frequency-selective fat-saturated images [12].
Our results confirmed high correlation between non-contrast MRI findings and surgical/histopathological findings regarding perinephric, collecting system and vascular invasion. Additionally, lymph node involvement was closely correlated.
The use of non-contrast MRI in our study for this patient category proved many advantages. It has the best soft tissue contrast over CT and ultrasonography, which can make up for the absence of the contrast material administration and the value of mass enhancement, the detection of macroscopic fat by the frequency-selective fat-saturated images, the detection of microscopic fat by dropping signal in the opposed phase images, detection of fluid as high signal in T2 WI and detection of subacute hemorrhage or presence of protein rich content as high signal in T1 WI [18].
The limitations in our study were encountered in small sample size. Besides, the detection of the vascularity and pattern of enhancement of the masses, detection of small masses with signal intensity similar to that of the renal parenchyma and the differentiation between different RCC subtypes could not be significantly obtained in our study.

Conclusion
Recognition of the most important imaging features of solid renal masses may assist in their proper diagnosis and management. Non-contrast multiparametric MR imaging as a noninvasive imaging method provides critical information that can help in differentiation of the most common renal masses, including the common RCC subtypes and AMLs, and as a result may assist in selecting the most appropriate management and follow-up of these lesions.