Comparison of Artificial Intelligence Techniques to Evaluate Performance of a Classifier for Automatic Grading of Prostate Cancer From Digitized Histopathologic Images.

Importance: Proper evaluation of the performance of artificial intelligence techniques in the analysis of digitized medical images is paramount for the adoption of such techniques by the medical community and regulatory agencies. Objectives: To compare several cross-validation (CV) approaches to evaluate the performance of a classifier for automatic grading of prostate cancer in digitized histopathologic images and compare the performance of the classifier when trained using data from 1 expert and multiple experts. Design, Setting, and Participants: This quality improvement study used tissue microarray data (333 cores) from 231 patients who underwent radical prostatectomy at the Vancouver General Hospital between June 27, 1997, and June 7, 2011. Digitized images of tissue cores were annotated by 6 pathologists for 4 classes (benign and Gleason grades 3, 4, and 5) between December 12, 2016, and October 5, 2017. Patches of 192 µm2 were extracted from these images. There was no overlap between patches. A deep learning classifier based on convolutional neural networks was trained to predict a class label from among the 4 classes (benign and Gleason grades 3, 4, and 5) for each image patch. The classification performance was evaluated in leave-patches-out CV, leave-cores-out CV, and leave-patients-out 20-fold CV. The analysis was performed between November 15, 2018, and January 1, 2019. Main Outcomes and Measures: The classifier performance was evaluated by its accuracy, sensitivity, and specificity in detection of cancer (benign vs cancer) and in low-grade vs high-grade differentiation (Gleason grade 3 vs grades 4-5). The statistical significance analysis was performed using the McNemar test. The agreement level between pathologists and the classifier was quantified using a quadratic-weighted κ statistic. Results: On 333 tissue microarray cores from 231 participants with prostate cancer (mean [SD] age, 63.2 [6.3] years), 20-fold leave-patches-out CV resulted in mean (SD) accuracy of 97.8% (1.2%), sensitivity of 98.5% (1.0%), and specificity of 97.5% (1.2%) for classifying benign patches vs cancerous patches. By contrast, 20-fold leave-patients-out CV resulted in mean (SD) accuracy of 85.8% (4.3%), sensitivity of 86.3% (4.1%), and specificity of 85.5% (7.2%). Similarly, 20-fold leave-cores-out CV resulted in mean (SD) accuracy of 86.7% (3.7%), sensitivity of 87.2% (4.0%), and specificity of 87.7% (5.5%). Results of McNemar tests showed that the leave-patches-out CV accuracy, sensitivity, and specificity were significantly higher than those for both leave-patients-out CV and leave-cores-out CV. Similar results were observed for classifying low-grade cancer vs high-grade cancer. When trained on a single expert, the overall agreement in grading between pathologists and the classifier ranged from 0.38 to 0.58; when trained using the majority vote among all experts, it was 0.60. Conclusions and Relevance: Results of this study suggest that in prostate cancer classification from histopathologic images, patch-wise CV and single-expert training and evaluation may lead to a biased estimation of classifier's performance. To allow reproducibility and facilitate comparison between automatic classification methods, studies in the field should evaluate their performance using patient-based CV and multiexpert data. Some of these conclusions may be generalizable to other histopathologic applications and to other applications of machine learning in medicine.

Automatic grading of prostate cancer in digitized histopathology images: Learning from multiple experts.

Prostate cancer (PCa) is a heterogeneous disease that is manifested in a diverse range of histologic patterns and its grading is therefore associated with an inter-observer variability among pathologists, which may lead to an under- or over-treatment of patients. In this work, we develop a computer aided diagnosis system for automatic grading of PCa in digitized histopathology images using supervised learning methods. Our pipeline comprises extraction of multi-scale features that include glandular, cellular, and image-based features. A number of novel features are proposed based on intra- and inter-nuclei properties; these features are shown to be among the most important ones for classification. We train our classifiers on 333 tissue microarray (TMA) cores that were sampled from 231 radical prostatectomy patients and annotated in detail by six pathologists for different Gleason grades. We also demonstrate the TMA-trained classifier's performance on additional 230 whole-mount slides of 56 patients, independent of the training dataset, by examining the automatic grading on manually marked lesions and randomly sampled 10% of the benign tissue. For the first time, we incorporate a probabilistic approach for supervised learning by multiple experts to account for the inter-observer grading variability. Through cross-validation experiments, the overall grading agreement of the classifier with the pathologists was found to be an unweighted kappa of 0.51, while the overall agreements between each pathologist and the others ranged from 0.45 to 0.62. These results suggest that our classifier's performance is within the inter-observer grading variability levels across the pathologists in our study, which are also consistent with those reported in the literature.

Lymphoproliferative Disorders of the Gastrointestinal Tract.

CONTEXT: - The diagnosis of gastrointestinal lymphoproliferative disorders can be challenging because of the small size of biopsies and the wide spectrum of lymphoproliferative disorders that may be encountered. OBJECTIVE: - To review the spectrum of lymphoproliferative disorders involving the gastrointestinal tract, highlighting potential pitfalls. DATA SOURCES: - Peer-reviewed articles and clinical experience. CONCLUSIONS: - Interpretation of lymphoid infiltrates in gastrointestinal biopsies requires synthesis of morphologic, immunophenotypic, molecular genetic, and clinical information. Knowledge of indolent lymphoproliferative disorders that may mimic aggressive lymphomas will help in preventing misdiagnoses.

IgG4-related disease and lymphocyte-variant hypereosinophilic syndrome: A comparative case series.

OBJECTIVE: To compare the clinical and laboratory features of IgG4-related disease (IgG4-RD) and lymphocyte-variant hypereosinophilic syndrome (L-HES), two rare diseases that often present with lymphadenopathy, gastrointestinal symptoms, eosinophilia, and elevated immunoglobulins/IgE. METHOD: Comparative case series of 31 patients with IgG4-RD and 13 patients with L-HES. RESULTS: Peripheral blood eosinophilia was present in eight of 31 patients with IgG4-RD compared to 13 of 13 patients with L-HES (median eosinophils 0.4 vs 7.0 giga/L, P=.001) and 12 of 20 patients with IgG4-RD had increased serum IgE compared to eight of 13 patients with L-HES, P=.930. Twenty-seven of 30 patients with IgG4-RD had elevated serum IgG4 compared to five of 12 patients with L-HES (median IgG4 9.6 g/L vs 0.80 g/L, P=.002). Flow cytometry demonstrated an aberrant T-cell phenotype in 7 of 23 patients with IgG4-RD and 13 of 13 patients with L-HES (P<.001). T-cell clonality by PCR was positive in 12 of 23 patients with IgG4-RD vs 10 of 13 patients with L-HES (P=.143). Patients in both groups received corticosteroids as first-line therapy. For refractory disease in IgG4-RD, rituximab was the most common steroid-sparing agent, whereas in L-HES, it was pegylated interferon-α-2a. CONCLUSION: The overlapping features of these two diseases with divergent treatment options demonstrate the importance of familiarity with both entities to optimize diagnosis and treatment.

Post-radiation epithelioid angiosarcoma of the urinary bladder and prostate.

Angiosarcoma of the lower urinary tract is exceedingly rare. A minority of cases are associated with local radiotherapy. Epithelioid angiosarcoma is a variant of angiosarcoma composed of large rounded epithelioid endothelial cells that are positive for cytokeratin on immunostaining. There are only two cases of post-radiation epithelioid angiosarcoma reported in the urinary bladder, and none in the prostate gland. We report a case of epithelioid angiosarcoma involving the urinary bladder and prostate in a patient with a history of radiotherapy for prostatic adenocarcinoma. A brief review of literature regarding post-radiation epithelioid angiosarcomas in the lower urinary tract is discussed.

Outcomes in splenic marginal zone lymphoma: analysis of 107 patients treated in British Columbia.

Splenic marginal zone lymphoma (SMZL) accounts for less than 2% of all non-Hodgkin lymphomas. We identified 107 cases diagnosed with SMZL between 1985 and 2012 from the British Columbia Cancer Agency Centre for Lymphoid Cancer and Lymphoma Pathology Databases. Patient characteristics were: median age 67 years (range 30-88), male 40%, stage IV 98%, splenomegaly 93%, bone marrow involvement 96%, peripheral blood involvement 87%. As initial treatment, 52 underwent splenectomy (10 with chemotherapy), 38 chemotherapy alone (21 chemoimmunotherapy containing rituximab, 1 rituximab alone), two antivirals for hepatitis C, and 15 were only observed. The 10-year overall survival for first-line splenectomy versus chemotherapy was 61% and 42%, respectively [Hazard Ratio (HR) 0·48, 95% confidence interval (CI) 0·26-0·88, P = 0·017]. The 10-year failure-free survival (FFS) after first-line splenectomy vs chemotherapy was 39% and 14%, respectively (HR 0·48, 95% CI 0·28-0·80, P = 0·004). Among the 38 patients who received first-line chemotherapy, FFS was similar between those receiving rituximab (n = 22) and those who did not (n = 16) (HR 0·64, 95% CI 0·31-1·34, P = 0·238). Fifteen patients transformed to aggressive lymphoma with median time to transformation of 3·5 years (range 6 months to 12 years) and the 10-year transformation rate was 18%. In conclusion, splenectomy remains a reasonable treatment for patients with SMZL.

Expression analysis of kidney-specific cadherin in a wide spectrum of traditional and newly recognized renal epithelial neoplasms: diagnostic and histogenetic implications.

Kidney-specific cadherin (Ksp-cad) is a membrane-associated cell adhesion glycoprotein expressed by the distal nephron tubular cells in its later developmental stages. Chromophobe renal cell carcinoma and renal oncocytoma are reported to be variably positive for Ksp-cad with some studies suggesting a discriminatory role for Ksp-cad. Immunoreactivity in other tumors with granular eosinophilic cytoplasm including clear cell and papillary renal cell carcinomas needs to be clearly elucidated and its expression in emerging novel and other unusual renal epithelial neoplasm subtypes including tumors with uncertain histogenesis is not yet known. In this study, we performed a detailed immunohistochemical analysis for Ksp-cad in a broad range of 136 renal epithelial neoplasms. Reactivity with Ksp-cad was observed in the following tumors: chromophobe renal cell carcinoma [23/25 (92%), diffuse (>50% of tumor cells)] positivity and membranous characteristically accentuating the "plant cell-like" histomorphology of the typical (clear) type, renal oncocytoma [15/20 (75%), usually diffuse staining with predominantly membranous accentuation], papillary renal cell carcinoma [5/17 (29%) all focal to moderate, eosinophilic type or type 2-3/7 (43%), basophilic type or type 1-2/10 (20%)], Xp11 translocation carcinoma [1/4 (25%), diffuse positivity] and clear cell renal cell carcinoma [6/36 (17%) all focal, clear cell renal cell carcinoma with prominent eosinophilic cells 1/7 (14%)]. Immunoreactivity was higher when evaluating whole histologic sections than with tissue microarrays for both chromophobe renal cell carcinoma (100% vs. 60%) and renal oncocytoma (100% vs. 55%). No immunoreactivity was observed in mucinous tubular and spindle cell carcinomas (0/23), high-grade collecting duct carcinomas (of Bellini) (0/3), renal medullary carcinomas (0/2), and urothelial carcinomas (0/6). Our study documents the immunoreactivity of Ksp-cad in the range of contemporarily classified renal epithelial neoplasms. The findings argue against the use of Ksp-cad in differentiating chromophobe renal cell carcinoma and renal oncocytomas and further support their relationship to the distal nephron. Ksp-cad may be helpful in distinguishing these two tumor types from clear cell renal cell carcinoma with prominent eosinophilic cells particularly in cases with limited tissue samples (ie, needle core biopsy). In the similar diagnostic setting, caution must be exercised, however, in differentiating chromophobe renal cell carcinoma and renal oncocytoma from the eosinophilic variant of papillary renal cell carcinoma as moderate expression of Ksp-cad may be observed in papillary renal cell carcinoma. The histogenesis of mucinous tubular and spindle cell carcinoma remains debatable as this tumor does not express Ksp-cad, which is highly expressed normally in the thick ascending loop of Henle and the distal convoluted tubules. In conclusion, Ksp-cad is a useful tumor type associated marker for distinguishing chromophobe renal cell carcinoma and renal oncocytoma from the wide range of nonintercalated cell-related adult renal epithelial neoplasms; addition of this marker to a panel comprised of other histologic subtype-associated markers may greatly facilitate histologic subclassification of adult renal epithelial neoplasms.

Immunohistochemical analysis of mucinous tubular and spindle cell carcinoma and papillary renal cell carcinoma of the kidney: significant immunophenotypic overlap warrants diagnostic caution.

Mucinous tubular and spindle cell carcinoma, a rare, recently described distinctive subtype of renal cell carcinoma, may have some morphologic similarities to the more common papillary renal cell carcinoma, particularly the basophilic (type 1) tumors with prominent solid growth pattern. Tumor circumscription, compact tubular architecture, focal papillations, mucin production and foam cells (features seen in both papillary renal cell carcinoma and mucinous tubular and spindle cell carcinoma), as well as spindle cell morphology, have resulted in some cases sent to us in consultation with a question of possible sarcomatoid papillary renal cell carcinoma. In this study, tissue microarrays with triplicate samples each from 27 mucinous tubular and spindle cell carcinomas and 20 papillary renal cell carcinomas were created to simulate experience in renal biopsy specimens. From immunohistochemistry (IHC) data, published in the contemporary literature, a panel consisting of alpha-methylacyl-CoA racemase (AMACR), cytokeratin 7 (CK7), epithelial membrane antigen (EMA), renal cell carcinoma marker (RCC Ma), CD10, high molecular weight cytokeratin (HMWK), and c-kit was designed to test its utility in differential diagnosis. The immunoreactivity in mucinous tubular and spindle cell carcinoma was AMACR 93%, CK7 81%, EMA 95%, RCC Ma 7%, CD10 15%, HMWK 15%, and c-kit 5% and in papillary renal cell carcinoma was AMACR 95%, CK7 65%, EMA 88%, RCC Ma 25%, CD10 80%, HWMK 15%, and c-kit 18%. This largest study to date on IHC of mucinous tubular and spindle cell carcinoma dispels the specificity of AMACR for papillary renal cell carcinoma among the RCC subtypes. The histogenesis of mucinous tubular and spindle cell carcinoma from the distal nephron continues to be debatable, as our study showed the expression of the proximal convoluted tubule-related marker AMACR among these tumors. Thus, in tumors with predominant compact tubular growth and focal papillary architectures, careful attention to the presence of a low-grade spindle cell population may be helpful in the distinction of mucinous tubular and spindle cell carcinoma, as the key immunohistochemical stains for papillary renal cell carcinoma are also expressed in this subtype of renal cell carcinoma.

Analysis of multiple biomarkers shows that lymphoma-associated macrophage (LAM) content is an independent predictor of survival in follicular lymphoma (FL).

We studied the role of multiple biomarkers in determining outcome in follicular lymphoma (FL), concentrating in particular on the role of benign macrophages. The study group consisted of uniformly staged and treated patients with FL enrolled in a phase 2 trial between 1987 and 1993. All patients were younger than 61 years of age, had advanced-stage FL, and were treated with a multiagent chemotherapy regimen, BP-VACOP (bleomycin, cisplatin, etoposide, doxorubicin, cyclophosphamide, vincristine, and prednisone), followed by involved region radiation. The median follow-up of living patients was 12.5 years, and the median survival was 16.3 years. The International Prognostic Index (IPI) was predictive of overall survival (OS) (P = .003). Biopsy specimens from all cases were stained with an anti-CD68 antibody. Of the 99 evaluable patients with FL, 87 had less than 15 CD68+ macrophages/high-power field (hpf) (median, 7; range, 1-14) and 12 had more than 15 CD68+ macrophages/hpf (median, 20; range, 16-25) with a median OS of 16.3 vs 5.0 years, respectively (P < .001). A multivariate Cox model that included the IPI score, the histologic grade, and the lymphoma-associated macrophage (LAM) score, showed IPI and LAM to be independent predictors of OS (P = .009 and P = .004, respectively). The LAM content of FL predicts survival, and these data support a prominent role for nonneoplastic immune cells in the biology of FL.

Distribution of cytokeratins and vimentin in adult renal neoplasms and normal renal tissue: potential utility of a cytokeratin antibody panel in the differential diagnosis of renal tumors.

Adult renal epithelial neoplasms (RENs) comprise several distinct clinicopathologic entities with potential prognostic and therapeutic differences. Individual cases can show overlapping morphologic features, necessitating the use of ancillary methods. The purpose of this study was to determine the diagnostic utility of cytokeratin (CK) subtype expression pattern in a wide range of adult RENs. RENs (including clear cell [conventional] renal cell carcinoma (RCC), papillary RCC, chromophobe RCC, renal oncocytoma, collecting duct carcinoma (CDC), renal medullary carcinoma (RMC), urothelial carcinoma, metanephric adenoma (MA), tubulocystic carcinoma (TC) (also known as low-grade collecting duct carcinoma), and mucinous tubular and spindle cell carcinoma) were immunostained for CK subtypes (CK5/CK6, 7, 8, 13, 14, 17, 18, 19, 20), high molecular weight CKs 1, 5, 10, 14 (HMWCK), and vimentin (Vim). The expression pattern of normal kidney was also examined and correlated with RENs. Although there is some overlap, subtypes of RENs show distinctive CK expression profiles that may be useful in several differential diagnostic settings. Clear cell RCCs typically showed a restricted expression pattern of CK8, CK18 and Vim. Papillary RCCs typically expressed CK7, CK8, CK18, CK19, and Vim and could be distinguished from MA (CK7-). Chromophobe RCCs were typically CK7+, CK8+, CK18+, and Vim-, and could be distinguished from oncocytomas (typically CK7-). In oncocytomas, nonspecific staining of unblocked endogenous biotin is a potentially significant diagnostic pitfall. CDC, RMC, and TC demonstrated similar CK expression profiles (with the exception of HMWCK expression limited to CDC), supporting a close relationship between these entities. A panel of CK5/CK6, CK17, and Vim may be helpful in distinguishing CDC (typically CK5/CK6-, CK17-, Vim+) and urothelial carcinoma (typically CK5/CK6+, CK17+, Vim-). In conclusion, CK expression patterns may be helpful in several differential diagnostic situations when dealing with adult RENs.

An immunohistochemical approach to the differential diagnosis of renal tumors.

Renal neoplasms comprise several distinct clinicopathologic entities with potential prognostic and the rapeutic differences. Although careful morphologic examination using sections stained with hematoxylin and eosin will allow for the correct diagnosis in the majority of cases, there is sufficient overlap between several entities such that ancillary techniques may be necessary to arrive at the correct diagnosis. In routine diagnostic surgical pathology practice of renal tumors, immunohistochemistry is the foremost ancillary technique. Using an approach based on common histologic patterns (tumors with clear cytoplasm, granular cytoplasm, tubulopapillary architecture, spindle cell morphology, small round-cell morphology, and infiltrating poorly differentiated carcinoma), we will discuss the utility of immunohistochemistry in the differential diagnosis of renal neoplasms. In recent years, needle biopsies from renal masses are being increasingly performed. In these small biopsies, the entire range of cytoarchitectural features that are generally necessary to make a diagnosis may not be fully appreciated. Immunohistochemistry may be helpful in this setting to narrow the differential diagnosis or to arrive at a definitive diagnosis. Finally, the use of immunohistochemistry for the confirmation of metastatic renal cell carcinoma presenting at distant sites will be discussed. Panels of immunohistochemical stains are proposed for different settings, including renal cell carcinoma (RCC) marker, CD10, and vimentin to suggest renal origin of a metastatic tumor, and markers to aid in subclassification of RCC, including parvalbumin and c-kit for chromophobe RCC, and cytokeratin 7 and alpha-methyl-acyl-CoA racemase for papillary RCC.

Soluble interleukin-13Ralpha2 decoy receptor inhibits Hodgkin's lymphoma growth in vitro and in vivo.

Recent studies have demonstrated that the malignant Reed-Sternberg cells of Hodgkin's lymphoma (HL) secrete and are responsive to interleukin (IL)-13. We hypothesized that overexpression of a soluble IL-13 decoy receptor (sIL-13Ralpha2) via adenoviral-mediated gene transfer would inhibit IL-13-induced Reed-Sternberg cell proliferation. Western blot and ELISA analysis verified expression of sIL-13Ralpha2 in cell lysates and supernatants of AdsIL-13Ralpha2-transduced COS-7 cells. Treatment of two IL-13-responsive HL-derived cell lines, HDLM-2 and L-1236, with AdsIL-13Ralpha2-conditioned medium, resulted in the inhibition of cell proliferation, and down-regulated the phosphorylation of signal transducer and activator of transcription 6 (STAT6), an important mediator of IL-13 signaling. i.v. delivery of AdsIL-13Ralpha2 in NOD/SCID mice with s.c. implanted HDLM-2 cells delayed tumor onset and growth while enhancing survival compared with control mice. Intratumoral administration of AdsIL-13Ralpha2 led to the regression or stabilization of established tumors and was associated with diminished STAT6 phosphorylation. Our data demonstrate that AdsIL-13Ralpha2 can suppress HL growth in vitro and in vivo.

Expression of alpha-methylacyl-CoA racemase (P504S) in nephrogenic adenoma: a significant immunohistochemical pitfall compounding the differential diagnosis with prostatic adenocarcinoma.

Alpha-Methylacyl-CoA racemase (AMACR, P504S) has recently been shown to be a useful marker for the diagnosis of prostatic adenocarcinoma and a potential aid in its distinction from its many mimics, one of which is the benign lesion, nephrogenic adenoma (NA). The goal of this study was to assess the expression of AMACR in NA by immunohistochemistry, as well as other potentially useful markers, high-molecular-weight cytokeratin clone 34betaE12, p63, and prostate-specific antigen (PSA). AMACR was expressed in 4/4 NAs involving the prostatic urethra and underlying stroma, and in 3/16 NAs involving the bladder. The prostatic cases showed circumferential granular cytoplasmic AMACR expression of at least moderate intensity, in >75% of tubules in 3 cases and in <10% of tubules in the remaining case. The AMACR-positive cases in the bladder typically showed focal weak noncircumferential staining of the tubules and stronger staining of the cells lining the papillae. 34betaE12 staining was observed in 1/4 prostatic NAs and 4/16 bladder NAs, typically in a cytoplasmic pattern in a minority of cells. p63 and PSA were negative in all cases. Our data indicate that NA of the prostatic urethra commonly expresses AMACR and lacks basal cell-specific markers, making it not only a potential morphologic mimic of prostatic adenocarcinoma but also a significant immunohistochemical mimic as well. Awareness of NA as a significant pitfall in the diagnosis of prostatic adenocarcinoma and careful examination of hematoxylin and eosin-stained sections remains the key to the correct diagnosis, which can be supported by a negative PSA stain.

Infarcted adenomatoid tumor: a report of five cases of a facet of a benign neoplasm that may cause diagnostic difficulty.

We describe five cases in which adenomatoid tumors showed extensive necrosis, presumably due to infarction, and posed diagnostic difficulty. The tumors occurred in four males (three with epididymal tumors and one with an intratesticular tumor) and one female (with a parafallopian tube tumor) 35 to 44 years of age. Two of the men presented with acute scrotal pain simulating epididymitis, and two with a palpable mass. The parafallopian tube tumor was an incidental finding. The tumors were solitary, grossly well-circumscribed, uniformly solid masses that ranged in size from 1.1 to 3.5 cm. Microscopically, they were all characterized by central necrosis with pale mummified adenomatoid tumor identified at least focally but often overshadowed by nondescript necrotic tissue. Viable adenomatoid tumor was identified in all cases but was minor in amount in two of them. The necrosis was surrounded by a florid reactive process of fibroblasts and myofibroblasts that had plump nuclei often with prominent nucleoli, and occasional mitoses. Two of the epididymal cases had adjacent rete testis showing epithelial hyperplasia with hyaline globule formation. The microscopic appearance often suggested the possibility of a malignant neoplasm because of: 1) blurring of the normal relatively easily identifiable junction between adenomatoid tumor and adjacent tissue; 2) irregular pseudo-infiltration of fat by reactive tissue and adenomatoid tumor; 3) paucity of typical adenomatoid tumor due to the infarction and the fact that viable tumor usually showed a solid pattern; and 4) atypia of the associated reactive cells. This unemphasized feature of adenomatoid tumors may potentially lead to more aggressive therapy than warranted if it is not correctly interpreted.

The role of interleukin 13 in classical Hodgkin lymphoma.

The Reed-Sternberg (RS) cells of classical Hodgkin lymphoma (cHL) produce several cytokines, which are thought to account for the unique clinical and pathologic features of this disease. We previously identified interleukin (IL)-13 expression as a common feature of cHL and have studied the potential role of this cytokine as an autocrine growth factor for RS cells. IL-13 and the IL-13-specific receptor chain (IL-13R alpha1) are frequently expressed in cHL-derived cell lines and in RS cells from biopsies of cHL tissues. In contrast, IL-13 expression in non-Hodgkin lymphoma (NHL) is uncommon. Neutralization of IL-13 in cultures of cHL-derived cell lines HDLM-2 and L-1236 leads to a dose-dependent inhibition of proliferation, and is associated with increased apoptosis in L-1236 cells. IL-13 neutralization also decreased activation of signal transducer and activator of transcription (STAT)6, an important mediator of IL-13 function. Moreover, STAT6 is often activated in RS cells from primary tumor samples, implying that IL-13 signaling is occurring in these cells in vivo. This review will describe the biologic activities of IL-13 in the immune system, and summarize the evidence implicating IL-13 as an autocrine growth factor for RS cells in cHL. Finally, we will discuss the potential influence of IL-13 on the reactive inflammatory infiltrate that is characteristic of cHL.

The role of cytokines in classical Hodgkin lymphoma.

The clinical and pathologic features of classical Hodgkin lymphoma (cHL) reflect an abnormal immune response that is thought to be due to the elaboration of a variety of cytokines by the malignant Reed-Sternberg (RS) cells or surrounding tissues. The majority of cHL cases are characterized by expression of tumor necrosis factor receptor (TNFR) family members and their ligands, as well as an unbalanced production of Th2 cytokines and chemokines. Activation of TNFR members results in constitutive activation of nuclear factor-kappa B (NF-kappa B), a transcription factor important for the in vitro and in vivo growth of RS cell lines. The expression of Th2 cytokines and chemokines leads to the reactive infiltrate of eosinophils, Th2 cells, and fibroblasts characteristic of cHL, and can also contribute to a local suppression of Th1 cell-mediated cellular immune response. Another particularly important growth and survival factor for RS cell lines is the Th2 cytokine interleukin 13, which is also commonly expressed by primary RS cells. In approximately 40% of cHL cases, the presence of Epstein-Barr virus influences the Th1/Th2 balance toward the production of Th1 cytokines and chemokines, but this shift is apparently insufficient for the stimulation of an effective antitumor cell-mediated immune response. This review summarizes the current literature on cytokine expression by and activity on RS cell lines and primary cHL tissues, examines cytokine signaling pathways in RS cells, and discusses the role that cytokines play in the specific clinical and pathologic features of cHL.

Signal transducer and activator of transcription 6 is frequently activated in Hodgkin and Reed-Sternberg cells of Hodgkin lymphoma.

The unique clinicopathologic features of Hodgkin lymphoma (HL) are due to the multiple cytokines produced by its neoplastic cells, the Hodgkin and Reed-Sternberg (HRS) cells. Cytokine signaling is mediated through the signal transducer and activator of transcription (STAT) family of transcription factors. Immunoblotting and immunohistochemistry were used to examine cell lines and tissue sections derived from patients with HL and non-Hodgkin lymphoma (NHL) for expression of activated STAT proteins. Constitutive phosphorylation of STAT6 and STAT3 was common in HL. STAT6 was constitutively phosphorylated in 5 of 5 HL cell lines and in HRS cells from 25 of 32 (78%) classical HL cases. STAT3 was constitutively phosphorylated in 4 of 5 HL cell lines and in HRS cells from 27 of 31 (87%) classical HL cases. Only 4 of 24 NHL cases demonstrated constitutive STAT6 activation, whereas STAT3 activation was observed in 6 of 13 (46%) cases of B-cell NHL and 8 of 11 (73%) cases of T-cell NHL. Constitutive STAT5 phosphorylation was not a common feature of HL or NHL. STAT6 mediates signaling by interleukin 13 (IL-13), a cytokine frequently expressed by HRS cells. Antibody-mediated neutralization of IL-13 resulted in significant decreases in both cellular proliferation and levels of phosphorylated STAT6 of HL cell lines. In conclusion, constitutive STAT6 phosphorylation is a common and distinctive feature of HRS cells in classical HL, whereas STAT3 activation was regularly present in both HL and NHL. These results suggest that IL-13 signaling is largely responsible for the constitutive STAT6 activation observed in HRS cells and further implicate IL-13 as an important growth factor in classical HL.