Secondary Tumors of the Urinary System: An Imaging Conundrum.

The urinary system is frequently the site of percutaneous biopsies, and the kidneys are the most frequently biopsied organs. Adequate pre-procedural imaging and patient work-up is of utmost importance ( 44 ). The probability of a renal mass being a metastasis in patients with extrarenal malignancy is almost equal to the likelihood of its being a primary renal tumor ( 7 ). Since imaging features of urinary system metastases cannot exclude the possibility of a primary tumor of the urinary system, histopathologic examination of renal masses in patients with an extra-urinary malignancy is necessary in the management of and the decisions regarding the treatment options of these patients ( 3 7 10 40 ). The percutaneous biopsy of solid renal and perirenal lesions are generally straightforward, but cystic lesions may deserve further attention, as the biopsy yields may be lower in these lesions ( 45 ). Targeting the solid parts of patients' cystic lesions may be more helpful in the diagnosis of semisolid renal and perirenal lesions ( 44 ). Bladder lesions may also be diagnosed with percutaneous biopsy. This approach may be even more important in patients who have difficulties with retrograde endourological procedures due to urethral stenosis ( 46 ).

The urinary tract may be involved in metastatic spread from other primary cancers. Information on imaging features of secondary urinary tract tumors is very limited and mostly based on anecdotal case reports. A variety of tumors may metastasize to the urinary tract and may present, mostly, with vague and non-specific imaging features. Breast cancer, malignant melanoma, lung cancer, head and neck malignancies, thyroid tumors, colorectal and pancreatic cancers, bone and soft tissue sarcomas, and Merkel cell carcinoma may metastasize to the urinary tract ( 1 2 3 ). The imaging features of secondary urinary tract tumors have not been very well described in the relevant literature, the findings are almost always non-specific, and diagnosis is mostly based on either tissue diagnosis or clinical history. It is not unwise to consider the occurrence of secondary urinary tract tumors in patients with advanced malignancy, but it should also be noted that metastasis to the urinary system may also be the first sign of detectable distant metastasis from an unrelated source ( 1 ). The timely recognition of secondary urinary tract tumors with ultrasonography (US), computed tomography (CT), and magnetic resonance imaging (MRI) is of paramount importance. As the outcome and treatment approaches to primary and secondary urinary tumors may vary significantly, prompt detection and differentiation have the potential to dramatically affect the clinical management. While the main approach to primary tumors is potentially curative surgery, in patients with secondary urinary tumors, surgical intervention can be deferred or may be performed in selected patients.

Metastasis to the kidney is not uncommon, occurring between 7% and 12% of patients with overall cancers at post-mortem examination ( 4 5 ). The actual detection rate is lower in clinical settings than autopsy data, as the post-mortem studies also include microscopic metastatic involvement, which is generally beyond the resolution of cross-sectional imaging modalities ( 6 ). A recent study showed the imaging incidence of metastases to the kidney to be 0.9% ( 7 ). The detection of metastasis in the kidneys is usually a sign of advanced disease and almost always a grave prognosis. The primary site can be almost anywhere in the body, but the most common primary sites are lung, breast, skin (melanoma), and gastrointestinal tract, with lung being the most common ( 4 5 6 8 ). Although renal metastases from extrarenal malignancies usually present with bilateral involvement, unilateral solitary metastasis may also occur. Hematogenous renal metastases from lung, breast, gastrointestinal tract, and hematologic malignancies and malignant melanoma tend to give rise to bilateral, multiple renal masses, while lung cancer and melanoma may also metastasize via lymphangitic spread, which may result in unilateral perirenal and renal involvement ( 8 9 ). Despite being extremely rare, it should be noted that in a small percentage of oncologic patients with non-renal malignancy, kidneys may be the only site of metastasis ( 10 ). The detection of these patients with solitary metastases is extremely important as these patients may benefit from surgical intervention ( 11 ). These solitary metastases are generally clinically silent and discovered during follow-up studies; however, local symptoms like hematuria may also be detected in small percentage of patients ( 11 ). These metastatic deposits occur due to hematogenous spread and present as cortical nodules, which likely represent the entrapment of the metastatic cells in glomerular capillary tufts ( 12 ). As uroepithelium is not typically involved, hematuria, even in patients with large masses, is rare ( 13 ). The mean duration between the diagnosis of a primary extrarenal tumor and renal metastasis is highly variable, but the mean interval time has been reported to be 2.2 years ( 8 ). The data on the imaging findings of metastases to the kidneys is sparse. The distinction between primary kidney tumors and metastases to the kidneys is of crucial importance as the treatment approaches to these two different clinical entities may be almost completely different. Although US has been reported as a useful tool for diagnosing renal metastases, it has several limitations ( 8 14 ). Small lesions (1–2 cm) and lesions with similar echotexture to the background renal parenchyma may be hard to discern with US. When detected by US, they appear as homogenous hypoechoic masses without increased through transmission; however, heterogenous or echogenic metastases have also been reported ( Fig. 1 ) ( 8 14 ). CT is generally the most commonly utilized imaging tool in the imaging follow-up of the abdomen in patients with known malignancy. CT, with its well-known clinical robustness, can also evaluate other sites in the abdomen in addition to the kidneys and may give a better picture of the full extent of tumor load in the abdomen. The CT appearances of renal metastases are mostly non-specific ( 15 ). On contrast-enhanced CT, most metastases have been reported to appear as endophytic, ill-defined, small, solid lesions with homogeneous enhancement ( Fig. 2 ) ( 7 ). They are typically multifocal and do not distort the kidney contours. They typically enhance less than the background renal parenchyma and mostly appear as either isodense or slightly hypodense (10–40 Hounsfield units [HU]) on unenhanced CT studies ( Fig. 3 ) ( 16 ). After contrast agent administration, they tend to mildly enhance (5–15 HU) compared to non-contrast images ( 15 ). Multifocality and bilaterality are other helpful clues for diagnosis ( 17 ). Metastatic deposits are mostly completely surrounded by renal parenchyma and rarely appear as an exophytic lesion ( Fig. 4 ) ( 18 ). Calcification may also be detected within the renal metastases. The underlying reasons for calcification include either secretions from the tumor cells (mucoid and papillary carcinomas) or cellular differentiation (osteosarcoma and chondrosarcomas) ( 19 ). Patel et al. ( 7 ) suggested that possibility of renal metastasis in the setting of a renal mass increases when the extrarenal malignancy has a higher stage. In their study, solid nature and endophytic location were significantly higher in renal metastases than primary renal tumors in patients with extrarenal malignancy ( 7 ). Not much has been published about the imaging findings of renal metastases on MRI. These lesions are generally hypointense compared to background renal parenchyma on T1-weighted images (T1WI) while heterogeneously hyperintense on T2-weighted images (T2WI). They mostly hypoenhance after contrast injection. Diffusion-weighted imaging (DWI) may help detection and characterization of kidney lesions ( 20 ). Renal metastases present with increased signal intensity on DWIs and decreased signal on apparent diffusion coefficient maps secondary to restricted diffusion of water molecules ( Fig. 5 ) ( 21 ). Renal cell cancer (RCC) is the main differential diagnosis, and differentiation may not be possible, especially in patients with solitary metastases. The most common subtype of RCC is a clear cell variant that is characterized by its high vascularity, which is an uncommon feature in renal metastases. The relatively hypovascular variants of RCC, such as papillary type, are also well known, and differentiation from metastasis may be difficult or even impossible in some patients. Hypointensity of papillary type RCC may be helpful for differential diagnosis as the metastatic renal masses are typically hyperintense on T2WIs ( 9 ). The infiltration of renal veins and inferior vena cava is also a probable indicator of RCC rather than a metastasis. An infiltrative pattern similar to renal metastasis can also be detected in sarcomatoid RCC variants, primary uroepithelial tumors, and primary renal lymphoproliferative disorders ( 19 ). A comparison of imaging features of RCC and renal metastases is presented in Table 1 . Metastases to the renal collecting system occur less frequently than renal parenchymal metastases. Imaging differentiation between uroepithelial metastases and primary urothelial tumors may be difficult, as both may appear as diffuse urothelial thickening or as a fungating exophytic mass ( Fig. 6 ) ( 9 ).

Ureters are most commonly affected by tumor extension from adjacent organs, and these involvements should be differentiated from true hematogenous and lymphatic metastases. Transmural infiltration, or less commonly, diffuse infiltration of the local mucosa with or without muscular layer involvement, may be observed in cases of hematogenous and lymphatic metastases ( Fig. 8 ) ( 28 ). This true metastatic involvement of the ureters is an uncommon clinical situation and a very rare cause of ureteral obstruction ( 29 30 ). Since its first description in 1909, only around 400 cases have been reported ( 29 31 32 ). Autopsy studies have also revealed only rare occurrences of ureteral metastases, and in a series of 10233 consecutive autopsies, the incidence of ureteral metastasis was only 0.37% ( 33 ). As the ureters do not have a continuous longitudinal network of lymphatic and blood vessels, they are relatively resistant to metastasis by these two routes ( 34 ). Despite the fact that most of the ureteric metastases are asymptomatic and incidentally diagnosed at autopsy, urinary obstruction may also occur ( 1 35 36 ). The breast and gastrointestinal systems comprise half of the cases, followed by prostate and uterine cancers comprising 30–40%, with gastric and lung cancers making up the remaining cases ( 1 ). Ureters may be involved in partial or full thickness ( 37 38 ). As the periureteral adventitia is rich in blood vessels, this part of the ureter is generally the first layer involved ( 1 ). From an imaging standpoint, the differentiation of ureteral metastasis from primary urothelial tumors might be difficult or even impossible. In our clinical practice, primary uroepithelial tumors of the ureter typically involve a long segment and demonstrate multifocality, whereas secondary tumors usually affect a relatively short segment. Clinical history may also help differential diagnosis as some tumors are more prone to uroepithelial metastasis than others. The detection of asymmetric wall enhancement, wall thickening, and urinary obstruction are the expected imaging findings. Strictures after stone passage, distal ureteral stenosis due to tuberculosis, amyloidosis, and endometriosis should be considered in the differential diagnosis ( Fig. 9 ). Endoscopic evaluation of the ureters with histopathologic evaluation may be required in most of the cases.

Urethral metastases are extremely rare, but occur more frequently than primary tumors of the urethra. Metastasis to the urethra most commonly results from prostate, bladder, lung, and colorectal cancers ( 42 43 ). Urethral and penile metastases may result from hematogenous involvement, instrumental spread, retrograde lymphatic spread, and direct extension. Retrograde lymphatic spread was reported to be responsible especially for metastasis from colon cancers ( 42 ). Superficial US performed with a linear probe can detect metastasis in the urethra. Abundant secondary tumoral involvement of urethra presents as solid nodular or cylindrical mass with heterogeneous contrast-enhancement ( Fig. 12 ). Superior soft-tissue resolution makes MRI a preferred imaging modality in assessment of secondary tumors of the urethra in terms of location, size, and local extension. Differentiation between primary urethral cancers and secondary urethral extension of adjacent malignancies such as rectal, vaginal, and cervical tumors can be accomplished with MRI.

Secondary tumors of the urinary bladder are rare and comprise less than 2% of all bladder cancers ( 2 39 ). The most common primary sources are stomach, breast, and colon cancer and melanoma. Urinary bladder metastases from a primary RCC are rare, with fewer than 40 reported cases in the medical literature, with neck and trigone the most commonly involved sites ( 2 40 ). The metastases may be clinically silent or may present with hematuria. The patients diagnosed with bladder metastasis are generally at an advanced stage, and the prognosis is usually dismal; however, in very rare instances, the bladder may be the only metastatic site ( 41 ). Imaging mostly displays polypoid or non-polypoid diffuse asymmetric wall thickening ( Figs. 10 , 11 ). Proper imaging technique is important for both the diagnosis and adequate staging. Differentiation of bladder metastases from primary bladder tumors may be difficult from an imaging standpoint, and pathological diagnosis is almost always indicated.

Metastasis to the urinary system in patients with non-urinary system primary tumors is a relatively rare clinical phenomenon. Early diagnosis and evaluation of these subjects is important to obtain a satisfactory patient outcome. Given the ever-increasing role of cross-sectional imaging in the oncologic patient population, it would not be irrational to predict that this diagnosis will be more commonly made in the near future. All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Informed consent was obtained from all individual participants included in the study.