In penile and urethral cancers, imaging has come to play a crucial role in enhancing the precision of clinical staging and facilitating optimal surgical planning. Over the years, great improvements have occurred in imaging. High-resolution magnetic resonance imaging (MRI) now represents the gold standard for evaluating the primary tumor and its local extension. Lymphotropic nanoparticle-enhanced MRI, dynamic sentinel lymph node biopsy, and ultrasonography with fine-needle aspiration seem to be the superior modalities for detecting malignant regional lymph nodes. Positron emission tomography combined with computed tomography has shown great promise as a whole body screen for the detection of distant metastases. Ultimately, the ability of imaging to augment clinical evaluation and enhance the accuracy of staging penile and urethral cancers will translate into improved surgical decision making and overall superior patient outcomes.
Although penile and urethral carcinomas represent less than 1% of all malignancies in North America, imaging is of vital importance in their management, because therapeutic decisions depend on accurate determination of clinical stage.
Present staging systems
The 2010 TNM staging system proposed by the American Joint Committee on Cancer is the staging system of choice for penile and urethral cancers ( Table 1 ). For both sites, the T stage is defined by the depth of tumor invasion, the N stage by the number and/or size of cancer-affected lymph nodes, and the M stage by the presence of distant metastases. Unique to urethral cancer, a second substaging system exists for urothelial carcinoma (UC) of the prostatic urethra (see Table 1 ) because its histologic findings, prognosis, and management are distinct. Arriving at the correct clinical TNM stage involves integrating physical examination, imaging, and biopsy findings, a process that is not always obvious. For example, determining whether regional inguinal lymph nodes are enlarged and, if so, whether the cause is cancer or infection/inflammation can be difficult.
| Penile Cancer | Urethral Cancer | UC of the Prostate | |
|---|---|---|---|
| T stage | |||
| Tx | Tumor cannot be assessed | Tumor cannot be assessed | Tumor cannot be assessed |
| Tis | CIS | CIS | CIS of prostatic urethra; CIS of prostatic ducts |
| Ta | Noninvasive carcinoma | Noninvasive carcinoma | — |
| T1 | Lamina propria invasion | Lamina propria invasion | |
| T1a | Lamina propria invasion: low grade and no LVI | — | — |
| T1b | Lamina propria invasion: high grade or LVI present | — | — |
| T2 | Invasion of corpus spongiosum or corpus cavernosum | Invasion of spongiosum, prostate, or periurethral muscle | Invasion of spongiosum, prostate, or periurethral muscle |
| T3 | Invasion of urethra | Invasion of cavernosum, vagina, or bladder neck | Extraprostatic extension (cavernosum, bladder neck, prostate capsule) |
| T4 | Invasion of adjacent structures | Invasion of adjacent structures | Invasion of adjacent structures (eg, bladder) |
| N stage | |||
| N0 | All regional nodes are negative | All regional nodes are negative | All regional nodes are negative |
| N1 | Single positive inguinal node | Single positive node <2 cm | Single positive node <2 cm |
| N2 | Multiple positive inguinal nodes | Single positive node >2 cm or multiple nodes | Single positive node >2 cm or multiple nodes |
| N3 | Extranodal tumor extension or positive pelvic nodes | — | — |
| M stage | |||
| M0 | No metastases | No metastases | No metastases |
| M1 | Distant metastases | Distant metastases | Distant metastases |
Imaging of the primary tumor
Retrograde Urethrography and Voiding Cystourethrography
Before computed tomography (CT) and magnetic resonance imaging (MRI), retrograde urethrography (RUG) and voiding cystourethrography (VCUG) were considered the cornerstones for defining and staging urethral tumors. In men, RUG was commonly used for evaluation of the anterior urethra and VCUG was the preferred modality for the posterior aspect. In women, VCUG was the preferred modality. Common signs of urethral cancer on RUG include narrowing of the urethral lumen, irregular margins with extravasation, filling defects, and obstructive changes. Multiple small mucosal nodules seen after instrumentation or cystectomy may represent UC implants in the urethra. Double-contrast RUG, a technique combining air and contrast dye, may improve the visualization of small distal urethral lesions.
Although RUG and VCUG delineate intraluminal defects quite well, their capacity to stage tumors is limited because they are unable to determine the extent of local tissue invasion, such as tumor growth into the corpus spongiosum or cavernosum ( Fig. 1 A, B). If the periurethral tissues cannot be imaged then a T stage cannot be determined. RUG and VCUG may also prove technically challenging if a high degree of urethral obstruction is present. At present, cystourethroscopy has largely replaced RUG and VCUG in the evaluation of known urethral cancers because it allows for visualization and biopsy of lesions in one setting. However, RUG remains an important imaging modality for urethral stricture disease, and because roughly 50% of patients with urethral cancer have a history of urethral stricture, unusual findings on RUG obtained for urethral stricture should prompt evaluation for urethral cancer.
Cavernosography
The injection of contrast media directly into the corpora cavernosa may show filling defects suggestive of tumoral invasion. Raghavaiah preformed preoperative cavernosography on 10 patients with penile cancer and found that it successfully staged patients when compared with pathologic examination. Cavernosography has also identified multifocal filling defects consistent with metastases. At present, cavernosography is not an imaging modality that is recommended for the evaluation of penile carcinoma because it is limited in its ability to evaluate tumor beyond the corporal bodies, and other imaging modalities that are more precise are available.
Ultrasonography
Although significant variation exists, squamous cell carcinoma of the penis most commonly presents as a heterogenous hypoechoic lesion on ultrasonography (US). Agrawal and colleagues found that of 59 patients with penile cancer, 36% had hyperechoic lesions, 47% had hypoechoic lesions, and 17% had lesions of mixed echogenicity. No association was found between echogenicity and tumor morphology or grade. A major benefit of US is that it can identify specific penile tissue planes, including the tunica albuginea, corpus cavernosum, corpus spongiosum, and urethra. Delineation of tumor involvement with structures at the glans penis is more difficult. Horenblas and colleagues found that US was not able to differentiate between subepithelial and spongiosal invasion at the glans. Despite this shortcoming, US was found to enhance the accuracy of clinical staging of local tumor extension (T). In the 2004 European Association of Urology (EAU) guidelines on penile cancer, penile US was deemed the initial diagnostic modality of choice to assist physical examination in determining the depth of tumor penetration, particularly if cavernosal invasion is suspected.
Although US has not been evaluated as an imaging modality for urethral cancer, it has been used to evaluate urethral strictures. Some groups have found US superior to alternative imaging modalities for identifying the number, length, depth, and overall severity of strictures. US can also demonstrate involvement of the corpus spongiosum, something that RUG cannot reliably identify. Case reports also suggest that transvaginal US can be helpful in distinguishing urethral tumors from other urethral pathologic conditions such as urethral diverticulae. The authors have found pelvic US to correctly identify a large urethral tumor in a woman complaining of pelvic pain ( Fig. 2 A). Similarly, transrectal US has been used to stage and biopsy UCs of the prostatic urethra. Invasion of the prostatic stroma, ejaculatory ducts, or periprostatic tissue by a hypoechoic tumor can be seen on transrectal US.
Strengths of US include low cost, noninvasiveness, and the lack of ionizing radiation. Limitations include operator dependence, inability to assess tumor spread beyond the penis, and the inability to adequately stage lesions of the glans.
CT and Positron Emission Tomography
With the advent of helical scanning and multiplanar reconstruction, multidimensional high-resolution CT is now possible. However, modern high-resolution CT still does not aid in adequate visualization of penile tissue planes and is therefore not considered an imaging modality of choice for evaluating the T stage of penile and urethral cancers ( Fig. 3 ). More recently, the fusion of CT and positron emission tomographic (PET) scanning has allowed functional and anatomic imaging to be combined in a single modality. At present, there are very little data published regarding PET combined with CT (PET/CT) in penile or urethral cancer. One small study of 13 patients with penile cancer found that PET/CT had a sensitivity and specificity of 75% for detecting primary penile tumors. However, PET/CT remains no better than conventional CT in determining T stage because PET/CT continues to be similarly limited by poor soft tissue discrimination ( Fig. 4 A, B).
Magnetic Resonance Imaging
MRI is the most sensitive imaging modality for the local staging of penile and urethral carcinomas. Advantages of MRI include superior soft tissue contrast, better spatial resolution, and lack of ionizing radiation. In addition, MRI provides better assessment of the penile fascial planes than physical examination.
Normal male anatomy
The penis is composed of 3 cylindrical bodies: 2 dorsal corpora cavernosa and a ventral corpus spongiosum. The corpora cavernosa and corpus spongiosum have a signal intensity on T1-weighted images that is slightly higher than skeletal muscle. On T2-weighted images all 3 corporal bodies are hyperintense because of blood pooling. The penile bulb (dilated portion of corpus spongiosum where it attaches to the perineal membrane) has a higher T2 signal intensity than the corpora cavernosa because of differences in blood flow ( Fig. 5 ). The 3 corporal bodies enhance with gadolinium. The corpus spongiosum completely encases the penile portion of the male urethra. On T2-weighted images, urethral smooth muscle appears less intense than the spongiosum.
The corpora of the penis is surrounded by 3 fascial layers. Each corpus cavernosum is contained within a layer of tunica albuginea. Buck fascia then wraps around the tunica albuginea and splits ventrally to cover the corpus spongiosum also. Therefore, Buck fascia is the layer that holds all 3 corporal bodies together. The tunica albuginea is not distinguishable from Buck fascia on MRI, and these 2 fascial layers appear jointly as a single low-intensity rim of tissue surrounding the corporal bodies on T1- and T2-weighted images. External to Buck fascia is the dartos fascia, which is also hypointense. A small quantity of fat lies between the Buck and the dartos fascia and dorsally contains the deep dorsal vein of the penis and the dorsal penile arteries. This fat layer appears hyperintense on T2-weighted image. The cavernosal arteries of the penis are visible on T2-weighted axial images. They appear as 2 low-intensity foci within the dorsal medial third of the corpora cavernosa.
Normal female anatomy
The female urethra has a characteristic targetoid appearance on T2-weighted images because of concentric layers of different signal intensity. The skeletal muscle of the rhabdosphincter appears as a low-intensity outer ring of tissue ( Fig. 6 ). Moving inward, the next ring of tissue is the urethral smooth muscle, which has a higher signal intensity than the rhabdosphincter. The next inner layer represents the submucosa, which has a lower signal intensity. The innermost layer is the urethral mucosa, which is a bright hyperintense ring. The lumen of the urethra is demarcated as a small, dark, central focus.
Urethral tumors
Urethral tumors are typically best seen on sagittal images and present an obvious mass effect. Tumors tend to show a decreased signal intensity relative to that of the surrounding normal corporal tissue on T1- and T2-weighted images ( Fig. 7 ). In women, the characteristic targetoid appearance of the urethra often becomes disrupted (see Fig. 2 B). Malignant lesions usually enhance with gadolinium. The accuracy of MRI for staging urethral cancer has been reported to be approximately 75%. However, these studies reflect older technology, and newer applications have likely improved the accuracy of MRI. Still, MRI cannot reliably distinguish benign from malignant urethral tumors, and therefore, biopsy remains indicated for urethral masses.
Penile tumors
Penile tumors are best evaluated on T2-weighted images because of the superior contrast resolution between the hypointense fascial layers (Buck fascia and the tunica albuginea) and the adjacent hyperintense corporal bodies. Squamous cell carcinomas typically appear as hypointense lesions when compared with the adjacent corpora. Although penile cancers enhance with gadolinium, the normal corporal bodies enhance even more, a property that makes contrast enhancement less useful for assessing the local tumor.
On MRI, nonsquamous variants of penile cancer can appear different from squamous carcinomas (see Fig. 4 C). For example, melanoma appears hyperintense on T1- and T2-weighted images and enhances much more avidly with gadolinium. Rhabdomyosarcoma of the penis, a tumor more common in pediatric populations, has an intermediate signal intensity on T1-weighted images and a heterogenous high signal intensity on T2-weighted images. This tumor shows heterogeneous enhancement after contrast administration.
Intracavernosal alprostadil
Technical factors can affect the accuracy of MRI for penile and urethral tumors. For example, induction of an artificial erection with an intracavernosal injection of 10 μg of prostaglandin E 1 (alprostadil) may improve the ability of MRI to locally stage penile cancer. In the flaccid penis, conventional MRI may have difficulty in distinguishing tumors that are confined to the tunica albuginea from those that invade into the corporal bodies ( Fig. 8 ). The corporal bodies may have a low T2 signal intensity due to fibrosis or a transiently reduced blood flow when the penis is flaccid, which can cause the normally sharp interface between the corporal bodies and the tunica albuginea to degrade. The erect penis is a larger imaging target, and its increased blood flow leads to a significantly stronger cavernosal signal intensity and thereby greater contrast from the tunica albuginea. Several reports have suggested better staging accuracy for penile cancer when MRI was combined with intracavernosal alprostadil. Because of the potential for increased complications with intracavernosal alprostadil, some groups of investigators reserve its use for situations in which clinical uncertainty remains after imaging the flaccid penis.
Endoluminal coils
Conventional MRI of penile and urethral cancers includes the use of a surface radiofrequency coil that is placed over the pelvis. Radiofrequency coils may be positioned in strategic configurations, which are closer to the anatomic location of imaging interest, to improve the signal-to-noise ratio of MRI and thereby produce sharper images. With respect to perineal/pelvic MRI, coils have been developed for placement in the lumina of the rectum, vagina, and urethra. The use of these endoluminal coils has been reported to improve the ability of MRI to diagnose urethral disease. However, endoluminal coil technology is not standardized and is not available in most medical centers. In addition, endoluminal coil placement can be painful and may be not feasible in cases of large perineal masses.
Imaging of the primary tumor
Retrograde Urethrography and Voiding Cystourethrography
Before computed tomography (CT) and magnetic resonance imaging (MRI), retrograde urethrography (RUG) and voiding cystourethrography (VCUG) were considered the cornerstones for defining and staging urethral tumors. In men, RUG was commonly used for evaluation of the anterior urethra and VCUG was the preferred modality for the posterior aspect. In women, VCUG was the preferred modality. Common signs of urethral cancer on RUG include narrowing of the urethral lumen, irregular margins with extravasation, filling defects, and obstructive changes. Multiple small mucosal nodules seen after instrumentation or cystectomy may represent UC implants in the urethra. Double-contrast RUG, a technique combining air and contrast dye, may improve the visualization of small distal urethral lesions.
Although RUG and VCUG delineate intraluminal defects quite well, their capacity to stage tumors is limited because they are unable to determine the extent of local tissue invasion, such as tumor growth into the corpus spongiosum or cavernosum ( Fig. 1 A, B). If the periurethral tissues cannot be imaged then a T stage cannot be determined. RUG and VCUG may also prove technically challenging if a high degree of urethral obstruction is present. At present, cystourethroscopy has largely replaced RUG and VCUG in the evaluation of known urethral cancers because it allows for visualization and biopsy of lesions in one setting. However, RUG remains an important imaging modality for urethral stricture disease, and because roughly 50% of patients with urethral cancer have a history of urethral stricture, unusual findings on RUG obtained for urethral stricture should prompt evaluation for urethral cancer.
Cavernosography
The injection of contrast media directly into the corpora cavernosa may show filling defects suggestive of tumoral invasion. Raghavaiah preformed preoperative cavernosography on 10 patients with penile cancer and found that it successfully staged patients when compared with pathologic examination. Cavernosography has also identified multifocal filling defects consistent with metastases. At present, cavernosography is not an imaging modality that is recommended for the evaluation of penile carcinoma because it is limited in its ability to evaluate tumor beyond the corporal bodies, and other imaging modalities that are more precise are available.
Ultrasonography
Although significant variation exists, squamous cell carcinoma of the penis most commonly presents as a heterogenous hypoechoic lesion on ultrasonography (US). Agrawal and colleagues found that of 59 patients with penile cancer, 36% had hyperechoic lesions, 47% had hypoechoic lesions, and 17% had lesions of mixed echogenicity. No association was found between echogenicity and tumor morphology or grade. A major benefit of US is that it can identify specific penile tissue planes, including the tunica albuginea, corpus cavernosum, corpus spongiosum, and urethra. Delineation of tumor involvement with structures at the glans penis is more difficult. Horenblas and colleagues found that US was not able to differentiate between subepithelial and spongiosal invasion at the glans. Despite this shortcoming, US was found to enhance the accuracy of clinical staging of local tumor extension (T). In the 2004 European Association of Urology (EAU) guidelines on penile cancer, penile US was deemed the initial diagnostic modality of choice to assist physical examination in determining the depth of tumor penetration, particularly if cavernosal invasion is suspected.
Although US has not been evaluated as an imaging modality for urethral cancer, it has been used to evaluate urethral strictures. Some groups have found US superior to alternative imaging modalities for identifying the number, length, depth, and overall severity of strictures. US can also demonstrate involvement of the corpus spongiosum, something that RUG cannot reliably identify. Case reports also suggest that transvaginal US can be helpful in distinguishing urethral tumors from other urethral pathologic conditions such as urethral diverticulae. The authors have found pelvic US to correctly identify a large urethral tumor in a woman complaining of pelvic pain ( Fig. 2 A). Similarly, transrectal US has been used to stage and biopsy UCs of the prostatic urethra. Invasion of the prostatic stroma, ejaculatory ducts, or periprostatic tissue by a hypoechoic tumor can be seen on transrectal US.
