|Home | About | Journals | Submit | Contact Us | Français|
Lymphoepithelioma-like carcinoma (LELC) of the urinary bladder is often mixed with conventional transitional cell carcinoma and/or other histotypes. The pathologist’s determination of the morphologic purity of a given LELC at the biopsy stage is a clinically relevant endeavour, because there is some anecdotal evidence suggesting that pure or predominant LELC may be comparatively chemosensitive and have a favorable prognostic profile, which may potentially offer the possibility of effective therapy without bladder resection. The precise degree of cellular pleomorphism that is allowed in a pure LELC is unclear. We describe herein an otherwise conventional and pure LELC that showed, in a localized area that constituted approximately 25% of the overall tumor volume, a two to six fold variation in nuclear size, including multinucleated tumor cells. These pleomorphic areas were set in the same lymphoplasmacytic infiltrate as their conventional counterparts, and similarly displayed cellular syncytia. We performed a detailed immunophenotypic comparison between the conventional areas and the pleomorphic areas. No significant differences were found between the 2 areas in overall lymphoplasmacytic or histiocytic density, lymphocytic CD4/CD8 ratio, and lymphoplasmacytic kappa/lambda ratio. Similarly, both displayed similar qualitative and quantitative staining indices for p53, Ki67, cytokeratin AE1/AE3 and p16INKa. Scattered cells were cytoplasmically beta-catenin positive exclusively in the pleomorphic areas; however these cells were not notably larger than the cells in the conventional areas. Both components were immunohistochemically negative for HMB-45, CD1a, the estrogen receptor, Epstein-Barr virus, CD117, D2-40, CD56, cytokeratin 20 and chromogranin. Clinicopathologic analysis of a series of cases is required to establish if there is any significance to nuclear pleomorphism in LELC. However, the phenotypic similarity between the 2 areas in this case, the intimate admixture of the pleomorphic cells with the lymphoplasmacytic infiltrate, and their syncytial pattern of growth, all suggest that pure LELC may display marked nuclear pleomorphism, and that this finding may not, in of itself, be a valid basis for removing a case from the “pure” group.
Lymphoepithelioma was originally described, and has been most comprehensively characterized in the head and neck region, where it is considered a variant of undifferentiated non-keratinizing squamous cell carcinoma that displays a strong association with the Epstein Barr virus (EBV) infection [1, 2]. Lymphoepithelioma is characterized by a syncytial growth pattern of large carcinomatous cells with vesicular nuclei and prominent nucleoli that are intimately admixed with a notably inflammatory infiltrate . Carcinomas with these morphologic features have been described in a wide variety of anatomic locations, where they are generally designated “lymphoepithelioma-like carcinomas”. In the urinary tract, lymphoepithelioma-like carcinoma (LELC) has been described in the kidney , renal pelvis , ureter , penile skin , prostate gland , urethra and urinary bladder [3, 9–13]. Most LELC of the urinary tract originate in the urinary bladder, where approximately 80 additional cases [12, 13] have been reported since its original description in 1991 .
LELC are frequently admixed with conventional transitional cell carcinomas and/or other histotypes. In one series of 30 LELC, 28 of which were primary in the bladder, 43% of cases showed a non-LELC carcinomatous component, which included invasive urothelial carcinoma, invasive adenocarcinoma, and invasive squamous cell carcinoma . Amin et al  categorized the 11 cases they reported into 3 groups based on the presence or lack thereof of these other histotypes: pure LELC, “predominant” LELC in which the LELC constituted >50% of the tumor volume, and “focal” LELC in which the LELC constituted ≤50% of the tumor volume.
There is some evidence that suggests that these categorizations may have some prognostic significance [10, 11]. We describe herein an otherwise conventional and pure LELC that showed a localized area of notable cellular pleomorphism, and investigate the propriety of including this finding within the morphologic spectrum of pure LELC.
A 64-year-old Caucasian man presented with an initially intermittent, then constant hematuria of 5-weeks’ duration. Following a bladder washing in which the presence of “atypical epithelial cells” were reported, the patient underwent a rigid cystoscopic procedure. A 2.7 cm polypoid mass was noted to be present in the left lateral bladder wall. No other abnormalities were identified. An excisional biopsy of the entire exophytic component of the mass was performed. Because tumor myoinvasion was equivocal after histopathologic analysis, and given the high rate of myoinvasion in this histotype, the patient underwent a second cystoscopic procedure 5 weeks later. Multiple biopsies of the tumor bed showed abundant detrusor muscle but no malignant cells. A whole body computed tomographic scan showed no evidence of metastatic disease A six-week course of intravesical Bacillus Calmette– Guérin was administered. Biopsies performed at the regular follow-up cystoscopic procedure 3 months later again showed no residual malignancy. Several options were presented to the patient, including cystectomy, chemotherapy, and observant follow-up, of which he elected to be closely followed. He shows no evidence of disease recurrence at last follow-up, 23.5 months after completion of his intravesical therapy.
The tumor was received in eight approximately equal sized pieces that measured 2.9 × 2.1 × 1.6 cm in aggregate. All pieces were routinely processed for microscopic examination. Six pieces showed the well characterized pathologic features of LELC of the urinary bladder. These included a proliferation of large but relatively monotonous cells with round vesicular nuclei, prominent nucleoli, and minimal cytoplasm that merged almost imperceptibly with those of the adjacent cells, resulting in a notably syncytial appearance (Figure 1a). Mitotic figures, including atypical forms, were numerous and confluent necrosis was absent. Numerous inflammatory cells, most of which were lymphocytes and plasma cells, were intimately admixed with the aforementioned cells (Figure 1a). Scattered histiocytic cells also formed part of the inflammatory infiltrate. These 6 fragments will henceforth be referred to as the “conventional” component of the LELC. Two of the fragments displayed cytologic features that were notably distinct from the others. Approximately one-quarter of these tumor fragments were comprised of conventional LELC with scattered cells that approximately 2 to 3 times the nuclear size of the conventional component (Figure 1b). Scattered multinucleated tumor giant cells were present in these areas (Figure 1c). In the other parts of the fragments, the cells displayed a prominent cellular pleomorphism, with cells showing an up to a six- fold variation in nuclear size (Figure 1d). The cells still retained the vesicular chromatin pattern of the conventional component, but a rim of amphophilic cytoplasm, which was occasionally abundant, was evident (Figures 1c and and1d).1d). A syncytial pattern of growth was also evident, as was conventional LELC cells, and the intimately admixed inflammatory component in the background (Figures 1b–1d). These 2 fragments will henceforth be referred to as the “pleomorphic” component of the LELC. There was no significant difference in mitotic index or inflammatory infiltrate density between the conventional and pleomorphic areas. No clear invasion of the detrusor muscle was identified, but tumor cells were identified immediately adjacent to the most superficial layer of the muscle. For immunohistochemistry, 5μm-thick sections were cut from representative paraffin embedded tissue blocks from both components, and were mounted on positively charged glass slides. After heating, deparaffinization, and rehydration of the tissue-mounted slides, immunohistochemical assays were implemented in an Axiom 36 autostainer (LabVision Corporation, Fremont, CA, USA). Negative and positive controls were assayed in parallel and showed appropriate patterns of reactivity. Immunohistochemical specifications and results are outlined in Table 1, and images from selected markers are shown in Figure 2. The epithelial cells in the conventional and pleomorphic components both displayed diffuse immunoreactivity for cytokeratin AE1/AE3 and p16 (Figure 2b). The ki67 labeling indices were similar for the epithelial components of both (approximately 60%). Similarly, the p53 staining index was similar for both (approximately 90%, Figures 2c and and2d).2d). Scattered cells (<1% of tumor volume) in the pleomorphic component displayed cytoplasmic positivity for beta-catenin (Figure 2a). However, these cells appeared to be some of the background conventional LELC cells in these areas. The larger cells were entirely negative for beta-catenin. The epithelial cells of both components were negative for CD138, CD4, CD8, estrogen receptor, Epstein-Barr virus, CD117, CD138, kappa/lambda, D2-40, chromogranin, CD56, cytokeratin 20 and HMB-45. Most of the inflammatory cells in both components were CD4 positive lymphocytes, and the overall CD4 to CD8 ratio for both components was similar (approximately 3:1). CD138 highlighted the scattered lymphoplasmacytic cells in each fragment. The kappa to lambda ratio in both components was similar (approximately 1:1). Scattered CD68 positive histiocytes were present in both components. As noted previously, the overall density of the inflammatory component was similar in the 2 components.
Urinary bladder carcinomas with a lymphoepithelioma-like component constitute less than 1% of all carcinomas encountered at this anatomic location [9, 10]. As such, there are no large systematically collected follow-up data on this histotype. However, there are some preliminary lines of evidence that suggest that pure or predominant LELC have a relatively favorable prognostic profile. In the study of Amin et al , the 3 patients with pure LELC were treated with transurethral resection of bladder tumor (TUR) and chemotherapy only. These 3 patients showed no evidence of disease at an average of 47.6 months follow-up. The 5 patients with predominant LELC also showed no evidence of disease at an average follow-up of 12.6 months. The 3 patients with a focal (≤50%) LELC were deceased at follow-up, 2 of which were directly tumor-related . Similar findings were reported by Holmang et al : 6 patients with pure or predominant LELC were alive at follow-up (13 months to 18 years), as compared with none of the patients with focal LELC . In the study of Lopez-Beltran et al , 3 patients with pure LELC were alive at last follow-up (21-47 months, average 33). Two of these 3 patients were treated with TUR and chemotherapy while the third was treated with a cystectomy . Sixty-six percent of the patients with predominant LELC were free of disease and 33% were dead of disease at follow-up. All 4 patients with focal LELC were dead of disease at follow-up . The comparatively favorable prognostic profile associated with pure or predominant LELC as was reported in the aforementioned studies were somewhat contradicted in a recent report from Tamas et al . In their series of 28 cases, which represents the largest reported thus far, LELC were classified as “pure” or “mixed”. For cases treated with cystectomy under this paradigm, no significant differences in recurrence-free survival were found between the pure and mixed groups. Two of three pure LELC treated with chemotherapy were free of disease at 4 and 65 months follow-up. The third patient, however, showed recurrent disease at 17 months. Two patients treated with TUR only showed no recurrent disease at 38 months and 48 months. Nevertheless, the preponderance of available evidence indicates that pure and predominant LELC is prognostically distinct from focal LELC. As such, in order to provide patients and clinicians with therapeutic options following a biopsy or TUR, the purity or lack thereof of a given LELC should be noted in a pathologic report.
The morphologic features of conventional LELC are well established. However, the full spectrum of changes that are permissible for a “pure” case is not entirely clear. Tamas et al  included 2 cases with “focal clear cell features” in their pure group. The authors also included in their pure group 2 cases with focal glandular differentiation . These glandular areas reportedly merged with the conventional LELC areas, and were similarly infiltrated heavily with lymphocytes .
The case described herein was an otherwise pure LELC that showed a minority population of highly pleomorphic cells. Although nuclear pleomorphism (in the range of a 2-fold variation in nuclear size) is characteristic of bladder LELC , we do not believe the degree of pleomorphism noted herein has been previously reported. These pleomorphic areas were set in the same lymphoplasmacytic infiltrate as their conventional counterparts, and similarly displayed cellular syncytia. We performed a detailed immunophenotypic comparison between the conventional areas and the pleomorphic areas. No significant differences were found between the 2 areas in overall lymphoplasmacytic or histiocytic density, lymphocytic CD4/CD8 ratio, and lymphoplasmacytic kappa/lambda ratio. Similarly, both displayed similar qualitative and quantitative staining indices for p16, cytokeratin AE1/AE3, p53 and Ki67. The significance of the scattered beta-catenin positive small cells in the pleomorphic areas is unclear. Based on their morphologic features and immunophenotype, we believe that the degree of cellular pleomorphism seen in this case should be included within the morphologic spectrum of pure LELC.
The differential diagnosis for this case includes chronic cystitis, malignant lymphoma, poorly differentiated carcinoma with a prominent stromal lymphocytic infiltrate, and the rare large cell neuroendocrine carcinoma with lymphoepithelioma-like features . The former 2 entities were not significant differential considerations, because the lymphoplasmacytic infiltrate was not sufficiently dense as to obscure the epithelial component, and the presence of an epithelial malignancy was apparent. The differential diagnostic question of LELC with a poorly differentiated, stromal lymphoplasmacyte-rich component versus pure LELC is essentially the subject of this report. Large cell neuroendocrine carcinoma with lymphoepithelioma-like features  was excluded due to the lack of immunoreactivity of the epithelial cells with antibodies to CD56 and chromogranin.
In summary, we have reported an unusually pleomorphic example of LELC. Clinicopathologic analysis of a series of cases is required to establish if there is any significance to nuclear pleomorphism in LELC. However, the phenotypic similarity between the 2 areas in this case, the intimate admixture of the pleomorphic cells with the lymphoplasmacytic infiltrate, and their synytial pattern of growth, all suggest that pure LELC may display marked nuclear pleomorphism, and that this finding may not, in of itself, be a valid basis for removing a case from the “pure” group.