In this article, we review prostatic lymphomas and discuss the differential diagnosis of high-grade malignant neoplasms of the prostate. We illustrate this with a case of a 46-year-old man seen with lower urinary tract obstruction who had diffuse involvement by a high-grade malignancy on prostate biopsy. Morphologic evaluation and immunohistochemistry were consistent with diffuse large B-cell lymphoma of the prostate. Workup with positron emission tomography-computed tomography demonstrated intensely hypermetabolic lymph nodes in the mediastinum, as well as widespread osseous involvement and involvement of the pancreatic tail, prostate, and urinary bladder, suggesting secondary prostatic involvement by a nodal lymphoma. Lymphomas of the prostate are uncommon in surgical pathology practice and usually represent secondary involvement from leukemia/lymphoma at a more typical site. Chronic lymphocytic leukemia/small lymphocytic lymphoma is the most common subtype.
A 46-year-old man was seen at the urology clinic with a several-year history of hematuria, dysuria, and obstructive symptoms, including urgency, hesitancy, incomplete voiding sensation, and weak urine stream. The symptoms had worsened substantially since onset. He also described pain in the lower back, hips, and groin, which had started several months before the current clinic visit. These had also worsened since onset. A review of systems was otherwise negative. He had no personal history of malignant neoplasm, sexually transmitted infection, or neurologic disease. Family history revealed melanoma in the patient's brother. Urine cultures obtained several months previously were negative. His serum prostate-specific antigen (PSA) level obtained 14 months previously was 1.6 ng/mL (normal is <4.0 ng/mL) (to convert PSA level to micrograms per liter, multiply by 1.0). Given these findings, he had been diagnosed as having benign prostatic hypertrophy and was treated with tamsulosin (an α-blocker) and dutasteride (a 5α-reductase inhibitor). The treatment did not improve symptoms significantly. He had undergone cystoscopy 6 months before the current clinic visit, which showed no evidence of tumor. Multiple episodes of gross hematuria had ensued as a complication of cystoscopy. Prostate ultrasonographic volumes 6 months and 1 month previously were estimated at 122 mL and 169 mL, respectively (normal is 20–30 mL). On physical examination at the current visit, the prostate was diffusely enlarged, firm, and irregular, with a nodule on the right lateral apex. Further workup showed that the serum PSA level had fallen to 0.8 ng/mL. An ultrasonography-guided prostate biopsy was recommended.
At the time of biopsy, the anterior transition zone was hypoechoic and irregular; 14 cores were obtained. At low magnification, there was diffuse infiltration of the prostate parenchyma by a cellular process, which effaced the prostatic parenchyma. Higher magnification revealed sheets of high-grade, discohesive tumor cells with large, vesicular nuclei and prominent nuclear contour irregularities. Prominent nucleoli were not present. Occasional mitotic figures were identified, and apoptotic debris was abundant. Cytoplasm was moderate in amount and clear to pale eosinophilic. There was substantial crush artifact. The tumor cells showed no glandular differentiated or evidence of cell cohesion. Scattered atrophic benign prostatic glands were occasionally scattered in the tumor cell infiltrate.
To characterize the neoplastic cells, a panel of immunostains was performed. The neoplastic cells were negative for cytokeratin cocktail, cytokeratin K903, p63, synaptophysin, chromogranin, and PSA. CD20 was diffusely positive. CD3 and CD5 highlighted background T cells. Cyclin D1 was performed to rule out mantle cell lymphoma and was negative. A diagnosis of diffuse large B-cell lymphoma (DLBCL) was rendered. Photomicrographs are shown in the Figure.
Subsequent positron emission tomography–computed tomography demonstrated intensely hypermetabolic lymph nodes in the mediastinum, as well as widespread osseous involvement and involvement of the pancreatic tail, prostate, and urinary bladder, suggesting secondary prostatic involvement by a nodal lymphoma. Staging bone marrow showed no involvement by lymphoma, and cerebrospinal fluid cytology showed no evidence of lymphoma. The patient was given 6 cycles of combined rituximab, cyclophosphamide, hydroxydaunorubicin, vincristine, and prednisone chemotherapy and intravesicular methotrexate. Repeat prostate biopsy showed no evidence of lymphoma. The patient was alive at the second follow-up year.
Lymphomas of the prostate are uncommonly encountered in surgical pathology practice. Two large consecutive series have shown that less than 1% of prostate specimens demonstrate leukemia or lymphoma. Specifically, 1092 consecutive prostatectomy specimens showed leukemia/lymphoma in 0.8% of cases in a study by Terris et al.1 Most were chronic lymphocytic leukemia/small lymphocytic lymphoma, with Hodgkin lymphoma and hairy cell leukemia representing the remaining cases. There were no cases of DLBCL. All cases also had leukemic/lymphomatous involvement in sampled pelvic lymph nodes. A similar study of 4831 consecutive prostatectomy specimens by Chu et al2 (including prostatectomy, transurethral resection, and needle biopsy specimens) showed leukemia/lymphoma in 0.3% (15 cases). Similar to the study by Terris et al, most (53%) were chronic lymphocytic leukemia/small lymphocytic lymphoma. Only 15% showed DLBCL, with the remaining being marginal zone, mantle cell, and follicular lymphomas. Based on these studies, prostatic leukemia/lymphoma is uncommon in standard surgical pathology practice, is typically low grade, and is frequently associated with concurrent pelvic lymph node involvement.
The distinction between primary lymphoma of the prostate and secondary spread from another site is often difficult. Criteria for primary involvement have been proposed by Bostwick and Mann,3 which include (1) symptoms attributable to prostatic enlargement, (2) the prostate as the predominant site of involvement, and (3) the absence of involvement of liver, spleen, or lymph nodes within 1 month of diagnosis. The largest study4 of primary prostatic lymphomas to date included only 22 cases pooled from 3 large tertiary referral centers. The study showed that primary prostatic lymphomas differ in subtype from those that involve the prostate secondarily, with most being DLBCL and with smaller numbers of chronic lymphocytic leukemia/small lymphocytic lymphoma, follicular lymphoma, and Burkitt lymphoma.
For both primary and secondary prostatic leukemia/lymphoma, the mean age at diagnosis is in the seventh decade.3,4 Primary prostatic lymphoma frequently causes lower urinary tract obstruction. Therefore, patients typically are seen with hesitancy, urgency, and weak urine stream. Hematuria is also a frequent finding. Digital rectal examination findings may mimic benign prostatic hypertrophy, with most cases demonstrating a diffusely enlarged or nodular prostate. However, the serum PSA level is typically not elevated.3 Secondary prostatic lymphoma is often discovered incidentally to prostate sampling performed under a suspicion for prostatic adenocarcinoma. Therefore, the initial findings are frequently those that raise the suspicion for prostate cancer such as an elevated serum PSA level. Secondary prostatic lymphoma may also manifest as obstructive symptoms,3 as were present in the case illustrated herein.
The differential diagnosis for prostatic tumors with the appearance of DLBCL is broad, and includes high-grade carcinomas (particularly prostatic adenocarcinoma and urothelial carcinoma), small cell carcinoma, sarcoma, and nonneoplastic processes such as granulomatous prostatitis and benign chronic inflammation with follicular hyperplasia. Workup with immunohistochemistry is frequently required (Table). High-grade prostatic adenocarcinomas (Gleason pattern 5) of the conventional type are often readily distinguishable from DLBCL because many cases demonstrate nuclear monomorphism and cellular cohesion with the formation of large sheets of tumor cells. Prostatic adenocarcinomas with infiltrative single cells are more difficult to distinguish from DLBCL. Immunohistochemistry will allow for the distinction in most cases. Most prostatic adenocarcinomas will be positive for PSA, prostate-specific membrane antigen, and pancytokeratin5 and negative for CD45 and CD20, while DLBCL will have the opposite immunoprofile. Although evidence of a conventional carcinoma in the presence of a poorly differentiated malignant neoplasm is often taken as evidence of dedifferentiation of the conventional carcinoma, this principle should not be applied to the prostate in most cases. That is, prostatic adenocarcinoma is extremely common, thus increasing the probability that a prostate involved by one neoplasm (eg, lymphoma) will also be involved by a separate conventional prostatic adenocarcinoma. In keeping with this, many cases of prostatic lymphoma are seen together with prostatic adenocarcinoma.1,2 Therefore, in cases demonstrating a conventional prostatic adenocarcinoma with a second, distinct population of high-grade malignant cells, additional workup may be required to exclude a second malignant neoplasm.
Conventional high-grade urothelial carcinoma may be identified on prostate needle biopsy. Most urothelial carcinomas show cellular cohesion in the form of solid tumor nests, and many will show noninvasive carcinoma in the surface urothelium, thus allowing for a distinction from DLBCL. However, urothelial carcinomas demonstrate extreme morphologic diversity, with some composed of invasive, discohesive tumor cells, thus overlapping with DLBCL. Immunohistochemistry resolves the differential diagnosis in most cases. Most urothelial carcinomas will be positive for pancytokeratin, CK903/34βE12, p63, and GATA3,5–7 and they will be negative for CD45 and CD20, while most DLBCLs will have the opposite immunophenotype. A panel that includes CK903/34βE12, p63, GATA3, PSA, and prostate-specific membrane antigen is also useful in distinguishing prostatic adenocarcinoma from urothelial carcinoma.5–7 Most important, small cell carcinoma may arise from both the prostate and the urothelium and may demonstrate close morphologic overlap with DLBCL. Most small cell carcinomas at these sites represent evolution from a conventional prostatic adenocarcinoma or urothelial carcinoma and are therefore frequently seen in association with the conventional carcinomas. Metastatic small cell carcinoma from the lung is also an important differential diagnosis. A history of high-grade prostatic adenocarcinoma, urothelial carcinoma, or pulmonary small cell carcinoma is thus important information to obtain. Although small carcinomas from different sites are indistinguishable by conventional histology and immunohistochemistry (including TTF-1, which is not specific to lung), rearrangements in ERG detectable by fluorescence in situ hybridization are present in approximately 50% of conventional prostatic adenocarcinomas and prostatic small carcinomas and have not been reported in small cell carcinomas arising from other organs.8–12 Regardless of origin, small cell carcinoma characteristically shows predominantly sheetlike growth, with small blue cells that have scant cytoplasm, high nuclear to cytoplasmic ratio, salt-and-pepper chromatin, nuclear molding, single-cell or geographic necrosis, crush artifact, and brisk mitotic activity/apoptosis. Immunohistochemistry will distinguish small cell carcinoma from DLBCL in most cases because small cell carcinoma will usually express pancytokeratin (perinuclear dotlike staining), chromogranin, and synaptophysin and will lack expression of CD45 and CD20. Although rare, high-grade angiosarcoma may have an epithelioid appearance and vesicular chromatin, similar to DLBCL. Most cases will express CD31 and ERG, in contrast to DLBCL.
We recommend using a panel of immunohistochemical stains in dealing with any poorly differentiated neoplasm involving the prostate, particularly in the inpatient setting, where treatment may be pending definitive diagnosis. Aberrant expression of markers may be seen in almost all high-grade tumors, making reliance on a single marker dangerous. Therefore, in such cases we begin with a broad spectrum of antibodies depending on the clinical and morphologic findings, often including pancytokeratin, p63, CK903/34βE12, GATA3, PSA, prostate-specific membrane antigen, CD45, CD20, CD3, chromogranin, and synaptophysin. Further workup is then tailored based on the results of this panel.
Benign processes are also an important consideration. Granulomatous prostatitis may be primary or secondary to treatment of urothelial carcinoma with intravesicular BCG vaccine.13,14 It appears as a lobulocentric proliferation of epithelioid histiocytes, often with associated chronic inflammation and eosinophils. It may involve and destroy the prostatic parenchyma, similar to DLBCL. Necrosis may also be present in granulomatous prostatitis, especially in cases secondary to BCG vaccine treatment. Despite these overlapping features, the histiocytes of granulomatous prostatitis do not display the high degree of cytologic atypia seen in DLBCL and do not demonstrate increased mitotic figures and apoptotic debris. Although both processes may express CD45, granulomatous prostatitis is composed of CD68+/CD20− histiocytes,15 while DLBCL is composed of malignant cells with the B-cell phenotype CD68−/CD20+. Chronic inflammation with germinal center formation is rarely encountered in the prostate. Although germinal centers contain immunoblasts, which are reminiscent of the cells of DLBCL, reactive germinal centers are usually well circumscribed, with a mixture of small centrocytes and larger centroblasts, a polarized architecture, and well-developed mantle zones.
In summary, although lymphomatous involvement of the prostate is rare, the differential diagnosis of prostatic involvement by a high-grade malignancy is frequently encountered. Morphology, clinical history and presentation, imaging results, and immunohistochemistry/ancillary studies typically must be considered to reach the correct diagnosis and patient management.
Dr Tomlins was supported by a University of Michigan prostate cancer SPORE career development award, the Prostate Cancer Foundation, and a Stand Up To Cancer–Prostate Cancer Foundation Prostate Dream Team translational cancer research grant. Stand Up To Cancer is a program of the Entertainment Industry Foundation (Pasadena, California) administered by the American Association for Cancer Research (SU2C-AACR-DT0712).
The University of Michigan has been issued a patent on the detection of ETS gene fusions in prostate cancer, on which Dr Tomlins is a coinventor. The University of Michigan licensed the diagnostic field of use to Gen-Probe, Inc (San Diego, California), which sublicensed some rights to Ventana Medical Systems, Inc (Tucson, Arizona). Neither company had a role in data collection, interpretation, or analysis and did not participate in the study design or the decision to submit for publication. Dr Tomlins has received honoraria from and serves as a consultant to Ventana. The other authors have no relevant financial interest in the products or companies described in this article.
Presented in part at New Frontiers in Pathology: An Update for Practicing Pathologists; September 28, 2013; Ann Arbor, Michigan.