FOXP3-stained image analysis for follicular lymphoma: Optimal adaptive thresholding with maximal nucleus coverage

Immunohistochemical detection of FOXP3 antigen is a usable marker for detection of regulatory T lymphocytes (TR) in formalin fixed and paraffin embedded sections of different types of tumor tissue. TR plays a major role in homeostasis of normal immune systems where they prevent auto reactivity of the immune system towards the host. This beneficial effect of TR is frequently "hijacked" by malignant cells where tumor-infiltrating regulatory T cells are recruited by the malignant nuclei to inhibit the beneficial immune response of the host against the tumor cells. In the majority of human solid tumors, an increased number of tumor-infiltrating FOXP3 positive TR is associated with worse outcome. However, in follicular lymphoma (FL) the impact of the number and distribution of TR on the outcome still remains controversial. In this study, we present a novel method to detect and enumerate nuclei from FOXP3 stained images of FL biopsies. The proposed method defines a new adaptive thresholding procedure, namely the optimal adaptive thresholding (OAT) method, which aims to minimize under-segmented and over-segmented nuclei for coarse segmentation. Next, we integrate a parameter free elliptical arc and line segment detector (ELSD) as additional information to refine segmentation results and to split most of the merged nuclei. Finally, we utilize a state-of-the-art super-pixel method, Simple Linear Iterative Clustering (SLIC) to split the rest of the merged nuclei. Our dataset consists of 13 region-of-interest images containing 769 negative and 88 positive nuclei. Three expert pathologists evaluated the method and reported sensitivity values in detecting negative and positive nuclei ranging from 83-100% and 90-95%, and precision values of 98-100% and 99-100%, respectively. The proposed solution can be used to investigate the impact of FOXP3 positive nuclei on the outcome and prognosis in FL.

Automatic detection of follicular regions in H&E images using iterative shape index

Follicular Lymphoma (FL) accounts for 20-25% of non-Hodgkin lymphomas in the United States. The first step in grading FL is identifying follicles. Our paper discusses a novel technique to segment follicular regions in H&E stained images. The method is based on three successive steps: (1) region-based segmentation, (2) iterative shape index (concavity index) calculation, (3) and recursive watershed. A novel aspect of this method is the use of iterative Concavity Index (CI) to control the follicular splitting process in recursive watershed. CI takes into consideration the convex hull of the object and the closest area surrounding it. The mean Zijbendos similarity index (ZSI) final segmentation score on fifteen cases was 78.33%, with a standard deviation of 2.83.

International standardized approach for flow cytometric residual disease monitoring in chronic lymphocytic leukaemia

The eradication of minimal residual disease (MRD) in chronic lymphocytic leukaemia (CLL) predicts for improved outcome. However, the wide variety of MRD techniques makes it difficult to interpret and compare different clinical trials. Our aim was to develop a standardized flow cytometric CLL-MRD assay and compare it to real-time quantitative allele-specific oligonucleotide (RQ-ASO) Immunoglobulin heavy chain gene (IgH) polymerase chain reaction (PCR). Analysis of 728 paired blood and marrow samples demonstrated high concordance (87%) for patients off-therapy. Blood analysis was equally or more sensitive than marrow in 92% of samples but marrow analysis was necessary to detect MRD within 3 months of alemtuzumab therapy. Assessment of 50 CLL-specific antibody combinations identified three (CD5/CD19 with CD20/CD38, CD81/CD22 and CD79b/CD43) with low inter-laboratory variation and false-detection rates. Experienced operators demonstrated an accuracy of 95.7% (specificity 98.8%, sensitivity 91.1%) in 141 samples with 0.01-0.1% CLL. There was close correlation and 95% concordance with RQ-ASO IgH-PCR for detection of CLL above 0.01%. The proposed flow cytometry approach is applicable to all sample types and therapeutic regimes, and sufficiently rapid and sensitive to guide therapy to an MRD-negativity in real time. These techniques may be used as a tool for assessing response and comparing the efficacy of different therapeutic approaches.

Collagenous gastritis: a case report, morphologic evaluation, and review

Collagenous gastritis is rare; there are only four previous case reports. Histologic features seem to overlap with the other "collagenous enterocolitides"; however, pathologic criteria are not yet established for the diagnosis of collagenous gastritis. We describe an additional case of ostensible collagenous gastritis in a patient who initially presented with celiac sprue and subsequently developed colonic manifestations of mucosal ulcerative colitis. Endoscopic biopsies of the stomach revealed deposition of patchy, very thick bandlike subepithelial collagen in gastric antral mucosa, focal superficial epithelial degeneration, numerous intraepithelial lymphocytes, and a dense lamina propria lymphoplasmacytic infiltrate. Image analysis evaluation of gastric antral biopsies demonstrated a mean thickness of subepithelial collagen of 27.07 micron. Morphologic comparison was made with age-matched control groups of 10 patients who had normal gastric mucosal biopsies and 10 patients who had "chronic" gastritis, which revealed mean subepithelial collagen measures of 1.37 micron and 1.19 micron, respectively. We compared these morphologic findings with those of all previous case reports of collagenous gastritis and propose a pathologic definition based on the limited combined data. It seems that subepithelial collagen is dramatically thickened in reported cases of collagenous gastritis, with a cumulative mean measure of 36.9 micron. It is also apparent from this and previous reports that the thickened subepithelial collagen is accompanied by a chronic or chronic active gastritis and sometimes intraepithelial lymphocytes and surface epithelial damage. Recently described associations of lymphocytic gastritis, sprue, and lymphocytic colitis as well as collagenous and lymphocytic colitis suggest a common pathogenesis that empirically may include collagenous gastritis in the same disease spectrum. We propose that collagenous gastritis can be confidently identified by using analogous defined features of collagenous colitis: subepithelial collagen more than 10 micron in a patchy distribution, lamina propria lymphoplasmacytic infiltrates, intraepithelial lymphocytes, and surface epithelial damage. Collagenous gastritis also seems to have the same spectrum of associated clinical findings as collagenous colitis, including frequent coexistence of celiac sprue, watery diarrhea syndrome, and female predominance.

Synchronous occurrence of collagenous colitis and pseudomembranous colitis

Synchronous collagenous and pseudomembranous colitis has not been previously reported. A 73-year-old woman presented with chronic watery diarrhea and abdominal cramping of six weeks' duration. Biopsies of the colon revealed findings of collagenous colitis involving the endoscopically normal right colon, and superimposed collagenous and pseudomembranous colitis involving the rectosigmoid colon. Endoscopically, the left colon revealed discrete ulcerative plaques, and Clostridium difficile toxin A assay was positive. The patient partially responded to a three-week regimen of metronidazole, and symptoms resolved completely with subsequent steroid therapy. At follow-up endoscopy four months later, colon biopsies demonstrated persistence of subepithelial collagen but no pseudomembranes. The patient remained asymptomatic during this interval. Collagenous colitis has been reported in association with other inflammatory bowel diseases, including lymphocytic colitis, sprue and idiopathic inflammatory bowel disease. This unique association of collagenous colitis with an endotoxigenic inflammatory bowel disease is presented with a review of related disease features.

The sphingomyelin-ceramide pathway participates in cytokine regulation of C-reactive protein and serum amyloid A, but not alpha-fibrinogen

Maximal induction of the acute-phase proteins C-reactive protein (CRP) and serum amyloid A (SAA) in the human hepatoma cell line Hep3B requires the combination of interleukin (IL)-6 and IL-1. In contrast, IL-1 inhibits fibrinogen induction by IL-6. To explore the possible participation of the sphingomyelin-ceramide pathway in the transduction of cytokine effects, the role of this pathway in expression of CRP, SAA and alpha-fibrinogen was investigated. The cell-permeable ceramide analogues C2 and C6 each greatly potentiated induction of both CRP and SAA mRNA by IL-6+IL-1beta but did not affect the responses of alpha-fibrinogen to IL-6 or to IL-6+IL-1beta. The combination of IL-6+IL-1beta led to increased turnover of sphingomyelin in Hep3B cells. D609, an inhibitor of ceramide production by acidic but not neutral sphingomyelinases, substantially inhibited induction of CRP and SAA by IL-6+IL-1beta. The ability of C2 and C6 to potentiate the effects of cytokines suggests that the sphingomyelin-ceramide pathway participates in induction of CRP and SAA by IL-6+IL-1beta under these experimental conditions, most likely by transducing the effects of IL-1beta. C2 and C6 were unable to substitute for IL-1beta in enhancing IL-6 effects on CRP and SAA, consistent with other reports indicating that the sphingomyelin-ceramide pathway is only a single component of multiple necessary converging pathways for induction of many genes. In contrast, this pathway does not appear to participate in mediating the inhibitory effects of IL-1beta on fibrinogen induction by IL-6.

Identification of a C-reactive protein binding site in two hepatic carboxylesterases capable of retaining C-reactive protein within the endoplasmic reticulum

C-reactive protein (CRP) is normally synthesized by hepatocytes at relatively low rates and is retained within the endoplasmic reticulum (ER) via interaction with two carboxylesterases (termed gp60a and gp60b), which themselves are restricted to the ER by their COOH-terminal retention signals (HIEL and HTEL). During the acute phase response, an increase in CRP synthesis is accompanied by a decrease in its ER retention as a result of a decrease in the CRP binding affinity of gp60b. Our previous data indicated that the esterase active site, the CRP binding site, and the ER retention signal are functionally distinct. In the present studies, we have identified CRP-binding peptides produced by proteolytic cleavage of gp60a. The sequence shared by two CRP-binding peptides indicated that the CRP binding site of gp60a is contained within residues 477-499. These results were confirmed by expression of cDNAs coding for gp60a and b as bacterial fusion proteins. Fusion proteins containing the complete esterase COOH terminus bound CRP, whereas those truncated at residue 477 (or the homologous site in gp60b) did not. Based on the known crystal structures of three homologous hydrolases, the putative CRP-binding site of the gp60s is located on the surface and is physically distant from the esterase active site and the COOH-terminal ER retention signal.

C-reactive protein and serum amyloid A mRNA stability following induction by cytokines

We determined the effects of cytokine withdrawal on C-reactive protein (CRP) and serum amyloid A (SAA) mRNA abundance in Hep3B cells following 24 h of preinduction with interleukin 6 plus interleukin 1 beta. After cytokine withdrawal, CRP transcription rate rapidly fell to undetectable levels and mRNA levels fell with a half-disappearance time of about 2.5 h. In view of the relatively small amount of CRP transcription occurring at this time, it is likely that this value closely reflects the actual half-life of CRP mRNA. In contrast, substantial SAA transcription continued for at least 8 h, while SAA mRNA fell with a half-disappearance time of about 8.5 h. It is not possible, under these conditions, to determine SAA mRNA half-life, but it clearly was no greater than 8.5 h. Both Actinomycin D (ActD) and cycloheximide enhanced the stability of SAA mRNA, strongly suggesting that SAA mRNA degradation requires synthesis of a short-lived protein. CRP mRNA stability was also enhanced by ActD, but cycloheximide did not have a protracted stabilizing effect, suggesting complex regulatory processes. These studies provide insight into the stability of CRP and SAA mRNA following induction with [IL-6 + IL-1 beta] and into the mechanisms regulating their degradation.

Induction of human serum amyloid A in Hep 3B cells by IL-6 and IL-1 beta involves both transcriptional and post-transcriptional mechanisms

Previous studies of murine serum amyloid A (SAA) regulation during inflammatory states or following exposure to macrophage-conditioned medium have raised the possibility that both post-transcriptional and transcriptional mechanisms participate in induction of this family of proteins. Since IL-6 and IL-1 have been shown to induce SAA in human hepatoma cell lines, we explored the possibility that these cytokines might induce human SAA through post-transcriptional as well as transcriptional mechanisms. In kinetic studies, we found that continuous exposure of Hep 3B cells to either IL-6 or IL-1 beta alone caused only minimal increases in SAA mRNA and marginal increases in transcription (as measured by nuclear runon). In contrast, the combination of these cytokines led to a 23-fold increase in transcription, maximal at 12 h, with continuing increase in mRNA, achieving levels more than 1,000-fold greater than baseline by 72 h. This massive disparity between increases in mRNA and in transcription rate strongly supports the participation of post-transcriptional mechanisms in SAA induction by (IL-6 + IL-1 beta), whereas the lag between peaks of transcription and mRNA abundance reflects a relatively slow degradation rate of SAA mRNA. As observed by other workers, mean size of SAA mRNA decreased progressively over the course of incubation. Simultaneous kinetic studies of complement factors B and C3, haptoglobin, and alpha-1 protease inhibitor revealed several different patterns of response to IL-6 and IL-1 beta.

Induction of human serum amyloid A in Hep 3B cells by IL-6 and IL-1 beta involves both transcriptional and post-transcriptional mechanisms

Previous studies of murine serum amyloid A (SAA) regulation during inflammatory states or following exposure to macrophage-conditioned medium have raised the possibility that both post-transcriptional and transcriptional mechanisms participate in induction of this family of proteins. Since IL-6 and IL-1 have been shown to induce SAA in human hepatoma cell lines, we explored the possibility that these cytokines might induce human SAA through post-transcriptional as well as transcriptional mechanisms. In kinetic studies, we found that continuous exposure of Hep 3B cells to either IL-6 or IL-1 beta alone caused only minimal increases in SAA mRNA and marginal increases in transcription (as measured by nuclear runon). In contrast, the combination of these cytokines led to a 23-fold increase in transcription, maximal at 12 h, with continuing increase in mRNA, achieving levels more than 1,000-fold greater than baseline by 72 h. This massive disparity between increases in mRNA and in transcription rate strongly supports the participation of post-transcriptional mechanisms in SAA induction by (IL-6 + IL-1 beta), whereas the lag between peaks of transcription and mRNA abundance reflects a relatively slow degradation rate of SAA mRNA. As observed by other workers, mean size of SAA mRNA decreased progressively over the course of incubation. Simultaneous kinetic studies of complement factors B and C3, haptoglobin, and alpha-1 protease inhibitor revealed several different patterns of response to IL-6 and IL-1 beta.

IL-1 receptor antagonist affects the plasma protein response of Hep 3B cells to conditioned medium from lipopolysaccharide-stimulated monocytes

The availability of the IL-1R antagonist (IL-1ra) has made it possible to assess the specific contributions of IL-1 to the acute phase changes induced by complex mixtures of cytokines. We utilized IL-1ra to define the contribution of IL-1 to the effects of conditioned medium from LPS-stimulated monocytes on production of the positive acute phase proteins C-reactive protein, serum amyloid A, fibrinogen, alpha 1-protease inhibitor, complement component C3, alpha 1-antichymotrypsin, alpha 1-acid glycoprotein, and ceruloplasmin and the negative acute phase proteins albumin and transferrin in Hep 3B cells. Induction of C-reactive protein and serum amyloid A was essentially abolished, induction of complement component C3 and alpha 1-acid glycoprotein was moderately decreased and induction of fibrinogen was enhanced. In contrast, there was no significant effect of IL-1ra on induction by conditioned medium of alpha 1-protease inhibitor, alpha 1-antichymotrypsin, or ceruloplasmin. IL-1ra partially blocked the down-regulatory effects of conditioned medium on both of the negative acute phase proteins we studied--albumin and transferrin. These findings enhance our understanding of the contribution of IL-1 to the acute phase response. In addition, they indicate that IL-1ra in vivo may influence synthesis of both positive and negative acute phase proteins.

IL-1 receptor antagonist affects the plasma protein response of Hep 3B cells to conditioned medium from lipopolysaccharide-stimulated monocytes

The availability of the IL-1R antagonist (IL-1ra) has made it possible to assess the specific contributions of IL-1 to the acute phase changes induced by complex mixtures of cytokines. We utilized IL-1ra to define the contribution of IL-1 to the effects of conditioned medium from LPS-stimulated monocytes on production of the positive acute phase proteins C-reactive protein, serum amyloid A, fibrinogen, alpha 1-protease inhibitor, complement component C3, alpha 1-antichymotrypsin, alpha 1-acid glycoprotein, and ceruloplasmin and the negative acute phase proteins albumin and transferrin in Hep 3B cells. Induction of C-reactive protein and serum amyloid A was essentially abolished, induction of complement component C3 and alpha 1-acid glycoprotein was moderately decreased and induction of fibrinogen was enhanced. In contrast, there was no significant effect of IL-1ra on induction by conditioned medium of alpha 1-protease inhibitor, alpha 1-antichymotrypsin, or ceruloplasmin. IL-1ra partially blocked the down-regulatory effects of conditioned medium on both of the negative acute phase proteins we studied--albumin and transferrin. These findings enhance our understanding of the contribution of IL-1 to the acute phase response. In addition, they indicate that IL-1ra in vivo may influence synthesis of both positive and negative acute phase proteins.

Induction of inflammatory cytokine release from cultured human monocytes by C-reactive protein

The human acute phase protein, C-reactive protein (CRP), is capable of specifically binding to and modulating the function of mononuclear phagocytes. To investigate whether CRP can also affect the capacity of these cells to produce inflammatory cytokines, enzyme immunoassays and Western blot techniques were used to quantitate interleukin 1 beta (IL-1 beta), interleukin 6 (IL-6) and tumor necrosis factor alpha (TNF-alpha) produced by freshly-isolated normal human monocytes. CRP induced the rapid release of each cytokine, with significantly elevated levels in culture supernatants at 4 hours and maximal levels of TNF-alpha at 8 hours, and of IL-1 beta and IL-6 at 16 hours of culture. The effects of CRP were dose-dependent; greater than 10-fold increases of each cytokine were observed following culture with greater than or equal to 50 micrograms/ml CRP, concentrations which are often found in the presence of moderate to severe inflammation or tissue injury. The induction of cytokine release by CRP was unaffected by inclusion of 25 micrograms/ml polymyxin-B in culture media, but was completely abrogated by prior boiling of the CRP, a procedure which had no effect on induction of monocyte cytokine release by lipopolysaccharide. The dose-dependent induction of inflammatory cytokines by CRP provides further support for the hypothesis that interaction with mononuclear phagocytes constitutes an important biological role for this acute phase protein.

Effect of flurbiprofen on cytokine production by human monocytes and U-937 and THP-1 cell lines

Possible effects of nonsteroidal antiinflammatory drugs (NSAID) on inflammatory mediators other than arachidonic acid metabolites which might contribute to the antiinflammatory effects of these drugs have not been fully explored. We investigated the effects of an NSAID, flurbiprofen, on production of the cytokines tumor necrosis factor alpha (TNF alpha), interleukin 1 beta (IL-1 beta) and interleukin 6 (IL-6) by human peripheral blood monocytes and by the human cell lines U-937 and THP-1. Cytokine production was induced by 1 microgram/ml bacterial lipopolysaccharide (LPS) in both monocytes and cell lines, and cytokine levels in supernatants were measured by enzyme immunoassay. In monocytes, IL-6 was the major product while in both cell lines, TNF alpha was the major product. Flurbiprofen caused moderate inhibition of IL-1 beta and TNF alpha production by stimulated monocytes, but did not affect IL-6 production. In contrast, flurbiprofen completely abolished IL-6 production by both cell lines and substantially inhibited IL-1 beta and TNF alpha production. These observations raise the possibility that inhibition of cytokine production by flurbiprofen may contribute to the antiinflammatory properties of this drug.

Decreased concentrations of tumor necrosis factor-alpha in supernatants of monocytes from homozygotes for hereditary hemochromatosis

To determine whether release of tumor necrosis factor-alpha (TNF-alpha), a cytokine that affects iron homeostasis, may be selectively altered in hereditary hemochromatosis, we measured concentrations of TNF-alpha and interleukin-1 beta (IL-1 beta) in supernatants of cultured peripheral blood monocytes from 11 homozygotes for hereditary hemochromatosis, 11 healthy individuals, and five patients with iron-loading anemia. The gene for hereditary hemochromatosis is tightly linked to the HLA locus on chromosome 6, but its exact site and product are not known. The gene for TNF-alpha also is located within the HLA region. Monocytes were incubated from 4 to 36 hours in medium alone or with added lipopolysaccharide. Mean concentrations of immunoreactive TNF-alpha in supernatants were significantly lower for subjects with hereditary hemochromatosis as compared to healthy controls (P less than .037) and patients with iron-loading anemia (P less than .005); differences between homozygotes for hemochromatosis and healthy controls were up to 4.5-fold at 4 hours (P = .008), 1.9-fold at 12 hours (P = .036), and 7.0-fold at 36 hours (P = .001). Importantly, concentrations of IL-1 beta in supernatants were not significantly different among the three groups. We conclude that release of TNF-alpha by monocytes may be selectively impaired in hereditary hemochromatosis. Deficient activity of TNF-alpha may contribute to the disordered iron metabolism of this disease.