新发与放疗后尿道腺癌：55 例临床病理及影像学研究De Novo and Postradiation Adenocarcinoma of the Urethra: A Clinicopathologic and Radiologic Study of 55 Cases.

Abstract

Herein, we report the results of clinicopathologic and radiographic analyses of 55 patients with primary adenocarcinoma of the urethra. Their ages at diagnosis ranged from 42 to 88 years (mean: 66 y). Twenty-six patients were male and 29 were female (ratio: 1.0:1.1). We noted 2 distinct groups of cases: de novo urethral adenocarcinoma (n=45) and urethral adenocarcinoma in patients after radiation therapy for prostatic adenocarcinoma (n=10), defined as postradiation adenocarcinoma. Postradiation tumors developed 8 to 25 years (mean: 15.6 y) after radiation therapy. Eight patients underwent brachytherapy. Among the 45 patients with de novo adenocarcinoma, 29 were female, for a female-to-male ratio of 1.8 to 1.0. In 17 female patients with de novo urethral adenocarcinoma, tumors developed in a urethral diverticulum. Histologically, urethral adenocarcinomas exhibited adenocarcinoma not otherwise specified or clear cell, mucinous, enteric, and adenoid cystic carcinoma morphology. In 2 cases of de novo adenocarcinoma and 3 cases of postradiation adenocarcinoma, the tumors progressed to sarcomatoid carcinoma. The mean disease-specific survival time for the entire cohort was 100.35 months. As reflected by the Kaplan-Meier curves for these patients, the patterns of disease-specific survival were similar in the de novo and postradiation groups.

Introduction

Primary adenocarcinoma of the urethra is a very rare tumor, with a slight predilection for female individuals. 1–4 Congenital diverticula and acquired strictures are predisposing factors contributing to the female predominance of urethral adenocarcinomas. 5 Because of the rarity of this tumor, most reports of it in the literature are case reports or small series, with clear cell adenocarcinoma being the most common reported subtype. 6–15 Comprehensive reports of larger series are exceedingly rare. 1,2 The clinical and morphologic spectrum of primary urethral adenocarcinoma remains incompletely defined, and its behavior is poorly understood. Therefore, in this retrospective study, we analyzed the clinicopathologic features of 55 cases of primary adenocarcinoma of the urethra, including 10 patients in whom tumors developed as a sequela of radiation therapy for prostatic adenocarcinoma.

Materials and Methods

Tissue Samples

The pathology archives of The University of Texas MD Anderson Cancer Center were searched for all primary urethral adenocarcinoma cases from January 1984 to April 2025. Pathology reports, hematoxylin- and eosin-stained slides, and pertinent immunohistochemical studies were reviewed. Primary urethral adenocarcinomas were defined as tumors arising in the urethra with pure glandular differentiation. Cases with secondary involvement of the urethra by urinary bladder adenocarcinoma or urothelial carcinoma with glandular differentiation were excluded. The cases were reviewed by 3 dedicated genitourinary pathologists (M.Z., P.T., and B.C.). Tumors were evaluated for gross appearance, size, in situ components, histologic features, adjacent organ/structure involvement, lymphovascular invasion, lymph node metastasis, and distant metastasis. The pathologic data were correlated with clinical data, including follow-up, treatment, and imaging data.

Immunohistochemical Staining

Immunohistochemical studies to detect the expression of the following were performed at UT MD Anderson: CK7 (Dako, Carpinteria, CA; diluted 1:100), p63 (Santa Cruz Biotechnology, Dallas, TX; diluted 1:1000), CK20 (Dako; diluted 1:40), thrombomodulin (Dako; diluted 1:40), CK5/6 (Dako; diluted 1:100), PIN dual cocktail (Biocare Medical, Pacheco, CA; diluted 1:40), CDX2 (BioGenex Laboratories, Fremont, CA; diluted 1:900), GATA-3 (Leica Microsystems, Wetzlar, Germany; diluted 1:200), SATB2 (Cell Marque, Rocklin, CA; ready to use), K27MH3.3 (Sigma-Aldrich, St. Louis, MO; diluted 1:500), NKX3.1 (Cell Marque; diluted 1:100), and PAX8 (Cell Marque; ready to use).

Human papillomavirus (HPV) in situ hybridization detection kit was used according to the manufacturer’s recommendations (Enzo, Long Island, NY; prediluted). Also, RNAscope analysis for HPV was carried out according to the manufacturer’s recommendations (Advanced Cell Diagnostics, Newark, CA). Briefly, formalin-fixed, paraffin-embedded tissue blocks were cut into sections at a mean±SD thickness of 5±1 μm. After deparaffinization and dehydration, HPV probes (HPV16 and HPV18), ubiquitin C, and bacterial gene dapB probes were added separately onto 3 adjacent tissue sections. Tissue sections were incubated with amplifiers and then with a 1:1 DAB mixture for signal detection. The sections were then counterstained and mounted.

Clinical Data

The study patients’ clinical information, including their age, sex, prior cancer history, and clinical/cystoscopy findings, was summarized. Follow-up information and patient outcomes were obtained from their medical records. The clinical presentations and treatments provided in the study cases were reviewed by the expert urologist (C.P.) subsequent to UT MD Anderson Institutional Review Board approval. The disease-specific survival (DSS) was analyzed, and Kaplan-Meier plots for de novo and postradiation groups were generated for comparison.

Radiologic Studies

Radiologic imaging findings were available for 35 patients, comprising 18 female and 17 male patients. Results of magnetic resonance imaging (MRI) studies were available for 28 patients (16 male, 12 female), whereas those of computed tomography (CT) studies were available for 7 patients (6 female, 1 male). MRI studies were performed at UT MD Anderson (n=13) or outside facilities (n=15). T2- and T1-weighted images from all MRI studies were available, with postcontrast sequences available for 24 patients.

Results

Clinical Presentation and History

We studied a total of 55 patients—29 female and 26 male (ratio: 1.1:1.0)—with a mean age at diagnosis of 66 years (range: 42 to 88 y). Their clinical, pathologic, and radiographic imaging data are summarized in Supplemental Table 1, Supplemental Digital Content 4, https://links.lww.com/PAS/C414 . Clinically, patients presented with urethral strictures, urinary incontinence, obstructive voiding symptoms, hematuria, vaginal itching and discharge, or perineal pain. On the basis of their clinical histories, 2 distinct groups of cases emerged: de novo urethral adenocarcinoma (45 patients) and urethral adenocarcinoma in patients with a history of radiation therapy for prostatic adenocarcinoma (10 patients), referred to as postradiation adenocarcinoma. In designating tumors as postradiation adenocarcinoma, we followed the Cahan criteria. 16 Among the 45 patients with de novo adenocarcinoma, 29 were female, for a female-to-male ratio of 1.8 to 1.0. In 17 of these female patients, tumors developed in a urethral diverticulum.

Postradiation urethral adenocarcinomas developed 8 to 25 years (mean: 15.6 y) after radiation therapy. Eight of these patients underwent brachytherapy, 1 underwent external beam radiation therapy, and 1 underwent an unidentified type of radiation therapy.

Gross Findings

Information on gross tumor features was available for 20 patients (de novo, n=18; postradiation, n=2) and was correlated with radiographic imaging ( Fig. 1A-G ). The mean tumor size was 3.2 cm (range: 0.5 to 6.2 cm). Typically, the tumors centered around the urethra and were friable, with poorly defined margins ( Fig. 1A-G ). The cut surface was yellow to grey and sometimes mucinous, with frequent occurrence of a coarse surface configuration. Tumor masses infiltrated adjacent structures, such as soft tissue surrounding the urethra and the prostate gland, bladder neck, vagina, or corpora spongiosa.

Figure 1.

Microscopic Findings

Histologically, the adenocarcinomas in this study were classified as various subtypes. In the de novo group, the subtypes were adenocarcinoma not otherwise specified (NOS; n=13) ( Fig. 2A, B ) and enteric (n=11) ( Fig. 2C, D ), mucinous (n=8) ( Fig. 3A-C ), clear cell (n=11) ( Fig. 3D, E ), and adenoid cystic (n=2) adenocarcinoma. In 2 cases of mucinous adenocarcinoma, the tumor had a prominent signet ring cell component. In another case, the tumor had 2 distinct components: enteric adenocarcinoma and large cell neuroendocrine carcinoma ( Fig. 4A-D ).

Figure 2.

Figure 3.

Figure 4.

In the de novo adenocarcinoma group, clear cell adenocarcinomas were positive for PAX8 ( Fig. 3E , inset). Nonclear cell adenocarcinomas had a distinctly different immunohistochemical profile, with CK7 (n=8), SATB2 (n=6), and CK20 (n=6) positivity in most cases. Strong diffuse nuclear CDX2 positivity was present in all cases tested using immunohistochemistry (n=8, Fig. 2D , inset). CK5/6 and p63 stains were negative (n=8). GATA-3 was negative (n=8). NKX3.1 was negative (n=20). All 3 cases subjected to HPV in situ hybridization (2 mucinous, 1 enteric) were negative for high-risk HPV. In the case of mixed enteric and large cell neuroendocrine carcinoma, the neuroendocrine component was strongly and diffusely positive for synaptophysin ( Fig. 4D , inset). Immunohistochemical stains for PSA, PSAP, and NKX3.1 were performed on this case to rule out the prostatic origin of the tumor and were negative on both components. In 2 cases of adenocarcinoma NOS, the tumor progressed to sarcomatoid carcinoma with features of high-grade undifferentiated spindle cell and pleomorphic sarcoma.

The postradiation adenocarcinomas consisted of adenocarcinoma NOS (n=4) ( Fig. 5A-D ) and enteric (n=2), enteric and mucinous (n=1), and mucinous (n=3) adenocarcinoma. In 3 of these cases, postradiation adenocarcinoma progressed to sarcomatoid carcinoma with features of high-grade undifferentiated spindle cell and pleomorphic sarcoma ( Fig. 6A-F ). In all cases of postradiation adenocarcinoma, the epithelioid components were positive for nuclear CDX2.

Figure 5.

Figure 6.

Loss of histone ME3H3 (H3K27) was evident in 2 cases of postradiation adenocarcinoma NOS, one of which progressed to sarcomatoid carcinoma, a phenomenon frequently seen in patients with postradiation sarcoma ( Figs. 5C-D, 6F ). In sarcomatoid postradiation adenocarcinoma cases, the loss of methylated histone ME3H3 (H3K27) was evident in both sarcomatoid and invasive adenocarcinoma components. The same case had retention of ME3H3 (H3K27) staining of the adenocarcinoma in situ within the adjacent overlying urethral mucosa.

We also observed that 2 de novo adenocarcinoma cases had adenocarcinoma in situ in the adjacent overlying urethral mucosa ( Fig. 7A-F ). In both of these cases, adenocarcinoma in situ developed on the background of glandular metaplasia, which comprised highly atypical glandular structures and multilayered glandular superficial proliferation.

Figure 7.

Radiologic Findings

Radiographic imaging data were available for 35 patients. In 28 patients in the de novo adenocarcinoma group, these data were available for the following histologic subtypes: enteric adenocarcinoma (n=7), mucinous adenocarcinoma (n=4), clear cell adenocarcinoma (n=7), adenocarcinoma NOS (n=8), and adenoid cystic carcinoma (n=2).

In female patients in the de novo adenocarcinoma group, more than half of the tumors developed in a urethral diverticulum ( Fig. 8A, B and Fig. S1A,B, Supplemental Digital Content 1, https://links.lww.com/PAS/C411 and Fig. S2A, Supplemental Digital Content 2, https://links.lww.com/PAS/C412 ). In 9 of the male patients, the tumors developed in the prostatic urethra. However, some of the cases also involved the bulbar/membranous urethra (n=8) as well as the penile urethra (n=6).

We noted significant overlap of imaging features among the urethral adenocarcinoma subtypes. Enteric adenocarcinomas typically presented as hypointense masses on T2-weighted magnetic resonance images, with avid early postcontrast enhancement and frequent invasion into adjacent structures. Adenocarcinoma NOS often demonstrated T2 hypointensity and strong postcontrast enhancement, with a tendency toward bulky or infiltrative morphology ( Fig. 8C, D ). Mucinous tumors exhibited variable signal characteristics—often hypointense on T2 images of male patients and hyperintense on those of female patients—with heterogeneous enhancement and diffusely infiltrative behavior ( Fig. 8A and Fig. S1C,D, Supplemental Digital Content 1, https://links.lww.com/PAS/C411 ). Clear cell carcinomas were characterized by hyperintense T2 signals and iso- to hyperintense T1 signals, demonstrating well-demarcated, expansile morphology with lobulated contours and septations, typically arising in diverticula ( Fig. 8A, B and Fig. S2A, Supplemental Digital Content 2, https://links.lww.com/PAS/C412 ).

In addition, clear cell carcinomas were the most morphologically distinct urethral adenocarcinomas, with lobulated, septated, well-demarcated masses (Fig. S1A,B, Supplemental Digital Content 1, https://links.lww.com/PAS/C411 and Fig. S2A, Supplemental Digital Content 2, https://links.lww.com/PAS/C412 ). They typically developed in female patients, but we had one example of clear cell adenocarcinoma developing in a male patient (Fig. S2C, Supplemental Digital Content 2, https://links.lww.com/PAS/C412 ). The enteric and NOS subtypes frequently exhibited more infiltrative growth patterns than the other subtypes did, with involvement of adjacent structures such as the prostate, corpus spongiosum, bladder, and vagina (Fig. S2B, Supplemental Digital Content 2, https://links.lww.com/PAS/C412 ). Mucinous tumors were often heterogeneous, with a diffusely infiltrative appearance and occasional calcifications (Fig. S1C,D, Supplemental Digital Content 1, https://links.lww.com/PAS/C411 ). A rare example of adenoid cystic carcinoma had nonspecific radiographic features with low signal intensity and developed in the area of the membranous male urethra (Fig. S2D, Supplemental Digital Content 2, https://links.lww.com/PAS/C412 ).

Imaging studies were available for 7 of the 10 patients with postradiation adenocarcinoma, comprising the following histologic subtypes: enteric adenocarcinoma (n=2), mucinous adenocarcinoma (n=1), enteric and mucinous adenocarcinoma (n=1), and adenocarcinoma NOS (n=3); 2 of the NOS cases progressed to sarcomatoid carcinoma. These 7 tumors demonstrated hypointense T2 signals, with invasion of the prostatic stroma and distortion of the prostatic anatomy, displacing brachytherapy seeds ( Fig. 8E and inset, Fig. S1E, Supplemental Digital Content 1, https://links.lww.com/PAS/C411 ).

Figure 8.

Treatment and Follow-Up

Information on metastatic spread was available for 12 patients (11 patients in the de novo group and 1 patient in the postradiation adenocarcinoma group). In the de novo adenocarcinoma group, 5 patients had metastases in pelvic lymph nodes only, whereas 4 patients had both lymph node and organ metastases. The metastatic spread in these 4 patients involved groin, inguinal, or pelvic lymph nodes combined with distant metastases in the liver, lung, vaginal wall, or skin. The 2 remaining patients had distant metastases only, which involved the liver or bone. In the postradiation group, one patient experienced pelvic lymph node metastasis.

Treatment information was available for 32 patients. In the de novo adenocarcinoma group, treatment information was available for 24 patients. Their treatments included chemotherapy before surgery (n=6), chemotherapy and radiation therapy (n=6), and chemotherapy only (n=4). In addition, 3 patients underwent surgery only, 2 had radiation therapy followed by surgery, and 1 underwent radiation therapy only. Furthermore, 1 patient underwent chemotherapy and radiation therapy followed by surgery. The remaining patient underwent chemotherapy, radiation therapy, and surgery followed by immunotherapy. In the postradiation group, treatment information was available for 8 patients. Two patients underwent surgery only, 2 underwent chemotherapy followed by surgery, 1 had chemotherapy only, 1 underwent surgery followed by chemotherapy, 1 had concurrent chemotherapy and radiation therapy, and 1 had neoadjuvant chemoimmunotherapy.

Follow-up information was available for 51 patients. In the de novo adenocarcinoma group, follow-up information was available for 41 patients (mean follow-up time: 111 mo; median follow-up time: 65 mo). In this group, at the end of follow-up, 14 patients were alive with no evidence of disease, 9 were alive with disease, 10 were dead of disease, 1 was dead of another cause, 1 was alive with an unknown disease status, and 6 were dead of unknown causes. In the postradiation adenocarcinoma group, follow-up information was available for all 10 patients (mean follow-up time: 43 mo; median follow-up time: 32 mo). Two patients were alive without disease, 5 were alive with disease, 2 died of disease, and 1 was alive with an unknown disease status.

The mean disease-specific survival duration for the entire cohort was 100.35 months, and the median duration was 55.53 months. The mean and median DSS times in the de novo adenocarcinoma group were 115.72 and 74.14 months, respectively. The mean and median DSS times in the postradiation adenocarcinoma were shorter (40.29 and 29.43 mo, respectively), but the differences were not significant. The factors that likely contributed to the lack of a significant difference included 7 deaths from unknown or other causes in the de novo group and the fact that the time-to-death patterns for the DSS were similar in the 2 groups as reflected in Kaplan-Meier survival curves (Fig. S3, Supplemental Digital Content 3, https://links.lww.com/PAS/C413 ). Consequently, the timing of DSS was similar for the entire cohort (mean: 31.28 mo; median: 26.53 mo) and the de novo (mean: 32.60 mo; median: 26.53 mo) and postradiation (mean: 24.68 mo; median: 24.68 mo) groups. We also analyzed survival in relation to specific pathologic subtypes of urethral adenocarcinomas and found that the differences among the groups were statistically not significant (Supplemental Table S2, Supplemental Digital Content 5, https://links.lww.com/PAS/C415 ). We performed multiple pairwise comparisons among the groups, and P -values ranged from 0.3 to 0.9. In summary, the follow-up data do not support that the clinical aggressiveness of de novo and postradiation urethral adenocarcinomas, as well as those of different pathologic subgroups, differ significantly.

Discussion

Primary urethral adenocarcinomas are exceedingly rare and affect both female and male individuals, with a slight predilection for female patients, as evidenced by the female-to-male ratio of 1.1:1.0 in our study. Most reports of these tumors in the literature are single case presentations or analyses of small series. 6–15 Our analysis of clinical, pathologic, and radiographic data identified 2 distinct groups of urethral adenocarcinoma developing de novo or after radiation therapy for prostatic adenocarcinoma. The de novo group had a clear preponderance of female patients, with a female-to-male ratio of 1.8:1.0. In this group, 58% of the female patients experienced tumor development in association with a urethral diverticulum. The association of adenocarcinoma with a diverticulum in the female urethra is well established in the literature. 17–21 In contrast, postradiation urethral adenocarcinomas developed exclusively in male patients as a sequela of radiation therapy for prostatic adenocarcinoma, which in the majority of cases was brachytherapy with prostatic radioactive seeds. Typically, these tumors developed many years after radiation therapy (mean: 15 y). In several prior publications, authors reported postradiation urethral adenocarcinoma virtually always as a sequela of brachytherapy for prostate cancer. 22–25 To the best of our knowledge, this is the first comprehensive report on postradiation urethral carcinoma.

Microscopically, adenocarcinomas of the urethra are classified as clear cell or nonclear cell tumors. Virtually all clear cell adenocarcinomas in our cohort affected female patients, and some of them developed in patients with pre-existing urethral diverticula. The nonclear cell types included enteric, NOS, and mucinous adenocarcinoma, with rare cases of adenoid cystic carcinoma. Some cases of urethral adenocarcinoma progressed to sarcomatoid carcinoma with features of high-grade spindle cell and pleomorphic sarcoma. Sarcomatoid carcinomas developing in the postradiation setting exhibited loss of histone H3K27, a hallmark feature of malignant peripheral nerve sheath tumors and a feature also seen in postradiation sarcoma cases. 26,27

The differential diagnosis included prostatic adenocarcinoma in male patients and potential involvement by tumors from the neighboring organs, or even metastasis. A rare solitary metastasis of gastrointestinal origin with enteric and mucinous features must be excluded by combined clinical, radiographic, and pathologic analyses. 28–31 Differential pathologic diagnosis of urethral adenocarcinoma is complex and requires exclusion of diverse entities, which may have overlapping microscopic features and clinical presentation. Secondary involvement of urothelial carcinoma of the bladder with glandular differentiation is a consideration, but it can be excluded by the analysis of clinical presentation and the retention of the urothelial phenotype in glandular structures with GATA3 positivity microscopically. Rare pure adenocarcinomas of the bladder with enteric and mucinous features may have overlapping microscopic features and immunophenotype but can be excluded by the combined analyses of clinical, cystoscopic, and radiographic presentations. Colorectal and gynecologic secondaries can be ruled out by the correlation with clinical history and radiographic presentation. In male patients, mucinous and ductal variants of prostatic adenocarcinoma can be excluded by negative immunohistochemical staining for prostate-specific markers. The differential diagnosis also includes a recently described primary urethral carcinoma, which had distinctive microscopic features and was positive for high-risk HPV16 and HPV18 in a majority of cases. 32 The presence of carcinoma in situ in the overlying mucosa provides the strongest support for urethral origin, but it is documented in a minority of cases.

Although radiographic imaging does not show specific features that can be correlated with histologic subtypes of urethral adenocarcinoma, it is of great help in identifying a tumor centered around the urethra, providing support for the urethral origin.

Clear cell adenocarcinomas of the urinary tract, including those developing in the urethra, resemble those that commonly develop in female genital organs. The histogenesis of these tumors is unclear, but Müllerian and mesonephric origins have been postulated. 33–37 Molecularly, these tumors frequently harbor mutations of the ARID1A and TP53 genes. 17 Authors have also reported mutations of PIK3CA and PTEN in clear cell adenocarcinoma cases. 38 In contrast, nonclear cell adenocarcinomas with enteric and mucinous features are characterized by frequent KRAS and EGFR mutations. 22

Available follow-up information demonstrates that urethral adenocarcinomas are highly aggressive, with mean and median timing of disease-specific death for our entire cohort of 31.28 and 26.53 months, respectively. The available data demonstrate there was no significant difference in the overall clinical behavior between de novo and postradiation urethral adenocarcinomas ( P =0.94). Similarly, there was no significant difference in clinical behavior among the various pathologic subtypes of urethral adenocarcinoma. However, we must acknowledge treatment heterogeneity and the limited number of cases available for the analysis as confounding variables.