A Narrative Review of Clinical Decision Support Systems for Perioperative Bleeding Management in Cardiac Surgery.

Bleeding complications in patients undergoing cardiac surgery are common. The clinician must assimilate multiple sources of monitoring information, make rational decisions on the etiology of the bleeding, and then formulate a treatment strategy. Clinical decision support systems that acquire this information and present the data in an easily usable format may be useful tools to guide the physician in optimizing treatment strategies through adherence to evidence-based best practice guidelines. The authors present a narrative review of the literature and discuss how clinical decision support systems may be useful to the clinician.

Effects of peri-operative bridging with low molecular weight heparins on coagulation during interruption of vitamin K antagonists: A mechanistic study.

BACKGROUND: Bridging with low molecular weight heparins (LMWH) is used in patients undergoing invasive procedures that require interruption of vitamin K antagonists (VKA). Little is known on the mechanisms underlying observed thrombotic and bleeding events. In this exploratory study we investigated the interactive effects of the co-administration of VKA, LMWH and surgery on coagulation. MATERIALS AND METHODS: Blood was sampled daily from day -3 to day +5 in 13 patients. In addition to measurement of INR, anti-Xa activity, thrombin generation (TG) testing and assessment of its protein determinants was performed. RESULTS: At the time of intervention the mean INR was 1.0 (SD 0.1, range 0.9-1.2); the mean residual anti-Xa level was 0.19 units/ml (SD 0.20 units/ml, range<0.05-0.60). The intervention caused a 2-3 fold increase in TG at day 0. Factor (F) XI had the strongest correlation with TG (peak and endogenous thrombin potential (ETP)) (r=0.6; p=0.02). Thrombomodulin-induced reduction of ETP increased from 10.0% (SD 9.2) at day -3 to 18.2% (SD 9.5) at day 0, p=0.02. After surgery, FVIII (175.9%(SD 58.9% to 246.7% (SD 71.4%); p=0.002) and fibrinogen (4.3 g/L (SD 1.1 g/L) to 5.6 g/L (SD 1.7 g/L); p=0.003) were significantly increased. CONCLUSIONS: Residual anti-Xa activity was present in 84.6% of patients at the day of the intervention. Three prothrombotic mechanisms were exposed: FXI dependent TG, reduced activity of the activated protein C pathway and postoperative rises in FVIII and fibrinogen. For the perioperative management the value of TG merits further study.

The effect of clinical decision support on adherence to thrombosis prophylaxis guidelines in medical patients; a single center experience.

INTRODUCTION: Venous thromboembolism (VTE) is an underestimated health problem. The administration of low molecular weight heparins (LMWH) to the appropriate patients dramatically decreases VTE incidence. Clinical decision support (CDS) might contribute to thrombosis prophylaxis guideline adherence. METHODS: A computerized integrated risk score program was used to estimate VTE and bleeding risk of nonsurgical patients. A VTE risk score of ≥4 resulted in an advice to administer LMWH. We selected 64 medical patients before the introduction of CDS (T0) and 64 patients after the introduction (T1). We compared guideline compliance between these groups using chi(2) tests. RESULTS: No difference between groups was found; Adherence to the guidelines at T0 was 59.4%, the same percentage of 59.4% was found at T1. To evaluate the effect of the introduction of CDS in terms of under and overtreatment we compared the prevalence of over and under treatment at T1 and T0. The OR for receiving under treatment at T1 compared to T0 is 0.48 (95% CI: 0.18-1.30), p = 0.14. The OR for overtreatment at T1 compared to T0 is 1.66 (95% CI: 0.74-3.73), p = 0.22 CONCLUSION: We found no improvement in guideline adherence towards anti thrombotic prophylaxis in medical patients after the introduction of CDS in this pilot study. There was however a non-significant shift towards over treatment. Possible explanations for these results are the increased awareness of the risk for thromboembolism induced by the study, suboptimal use of CDS and deviation from CDS advice caused by patient's preferences.

A comprehensive sensitivity analysis of microarray breast cancer classification under feature variability.

BACKGROUND: Large discrepancies in signature composition and outcome concordance have been observed between different microarray breast cancer expression profiling studies. This is often ascribed to differences in array platform as well as biological variability. We conjecture that other reasons for the observed discrepancies are the measurement error associated with each feature and the choice of preprocessing method. Microarray data are known to be subject to technical variation and the confidence intervals around individual point estimates of expression levels can be wide. Furthermore, the estimated expression values also vary depending on the selected preprocessing scheme. In microarray breast cancer classification studies, however, these two forms of feature variability are almost always ignored and hence their exact role is unclear. RESULTS: We have performed a comprehensive sensitivity analysis of microarray breast cancer classification under the two types of feature variability mentioned above. We used data from six state of the art preprocessing methods, using a compendium consisting of eight different datasets, involving 1131 hybridizations, containing data from both one and two-color array technology. For a wide range of classifiers, we performed a joint study on performance, concordance and stability. In the stability analysis we explicitly tested classifiers for their noise tolerance by using perturbed expression profiles that are based on uncertainty information directly related to the preprocessing methods. Our results indicate that signature composition is strongly influenced by feature variability, even if the array platform and the stratification of patient samples are identical. In addition, we show that there is often a high level of discordance between individual class assignments for signatures constructed on data coming from different preprocessing schemes, even if the actual signature composition is identical. CONCLUSION: Feature variability can have a strong impact on breast cancer signature composition, as well as the classification of individual patient samples. We therefore strongly recommend that feature variability is considered in analyzing data from microarray breast cancer expression profiling experiments.

Immunostaining of the androgen receptor and sequence analysis of its DNA-binding domain in canine prostate cancer.

Prostate cancer in the dog (cPC) has many features in common with hormone refractory human prostate cancer. As cPC is seen more often in castrated dogs, the contribution of the androgen receptor (AR) to the development of prostate cancer remains questionable. The aim of the present study was to evaluate the presence of the AR by immunohistochemistry in cPC. AR staining was observed in most tumors from intact and castrated dogs, but the proportion of positive cells and the staining intensity were much lower than in the prostate of healthy, non-castrated dogs. Most of the positive staining was seen in the cytoplasm rather than in the nuclei of the tumor cells. The predominant cytoplasmic localization was not related to mutations in exon 3 of the DNA-binding domain of the AR, as shown by sequence analysis of microdissected AR positive tumor cells. Other mechanisms that lead to an impaired androgen-AR signaling or a basal/stem cell like origin may explain the low cytoplasmic AR staining in cPC.

Comparative characterization of the canine normal prostate in intact and castrated animals.

BACKGROUND: Prostate diseases in the dog are generally regarded as representative for their human counterparts. We characterized the normal canine prostate in comparison to the normal human prostate. METHODS: Prostates of dogs were examined histomorphologically and by immunohistochemical detection of the markers CK14, HMWCK, CK5, CK18, CK7, UPIII, PSA, and PSMA. RESULTS: Histomorphologically, the canine prostate lacks the human zonal differentiation, has much more prominent acini, while comprising less stromal tissue. In general, the canine prostate epithelium displayed a highly differentiated character, with no cells expressing CK14, minimal amounts of cells expressing HMWCK/CK5 and the vast majority of cells expressing CK18 and PSA. After castration, the prostate epithelium regressed, and the remaining tubules were largely populated by cells showing a ductal phenotype (HMWCK+/CK5+/CK18+/CK7+). CONCLUSIONS: The human and canine prostate are histologically differently organized. The general scheme of cellular differentiation of the prostate epithelium may however be applicable to both species.

Histopathological and immunohistochemical characterization of canine prostate cancer.

BACKGROUND: In this study we try to identify the origin of canine prostate cancer (cPC) by classifying the tumors histological subtypes and relate these subtypes to their combined expressional characteristics of several tissue specific and differentiation markers. METHODS: cPCs were examined histomorphologically and by immunohistochemical detection of the cytokeratin markers CK14, HMWCK, CK5, CK18, and CK7, and of the markers UPIII, PSA and PSMA. RESULTS: Histopathologically, six growth patterns could be differentiated. The most frequent patterns were solid, cribriform and micropapillary growth patterns, while sarcomatoid, small acinar/ductal, and tubulo-papillary growth patterns were less frequent present. Solid growth patterns were significantly (P = 0.027) more often seen in castrated dogs. Immunohistochemically, about half of the cPC cases showed expression of PSA (8/20) and PSMA (10/20); 85% and 60% of the cPC expressed UPIII (17/20) and CK7 (12/20), while 13 and 12 cPC expressed CK5 and CK14, respectively; all cPC expressed CK18. CK14 was significantly more often and UPIII less frequent expressed in the solid growth patterns than in the micropapillary and cribriform patterns, respectively. CONCLUSIONS: Canine prostate cancer appear to be more aggressive and of a less differentiated type than most common human prostate cancers. Comparing the expression patterns of the markers in cPC to those in normal canine prostate tissue, cPC most likely originates from the collecting ducts rather than from the peripheral acini. Given also the fact that canine prostate cancer is unresponsive to androgen withdrawal therapy, canine prostate cancer mostly resembles human, androgen refractory, poorly differentiated prostate cancer.

Regulation of COX-2 expression in canine prostate carcinoma: increased COX-2 expression is not related to inflammation.

BACKGROUND: Cyclooxygenase-2 (COX-2) expression has been documented in human and canine prostate carcinoma (PCA). Canine PCA is a histologically heterogeneous tumor, sometimes including inflammatory infiltrates. However, it is unknown whether COX-2 expression in canine PCA is related to the histologic type of tumor, to the presence of inflammation, or to both. Moreover, little is known about the mechanisms regulating COX-2 expression in neoplastic tissue. HYPOTHESIS: COX-2 expression is related to the presence of inflammation in canine PCA and correlates with the degree of tumor differentiation. METHODS: The expression of COX-2 was examined in 28 cases of canine PCA by immunohistochemistry. In addition, a neoplastic and a nonneoplastic canine prostatic cell line were used to investigate the effects of interleukin-6 (IL-6), tumor necrosis factor-alpha (TNF-alpha), phorbol 12-myristate 13-acetate (PMA), epithelial growth factor (EGF), and specific signal transduction pathway inhibitors on COX-2 expression. RESULTS: Twenty-four of the 28 prostate tumors showed COX-2 expression. The presence of inflammatory infiltrates in tumor tissue was associated with lower COX-2 expression scores. In vitro, TNF-alpha, IL-6, and EGF increased COX-2 expression in nonneoplastic cells but not in PCA cells, where baseline expression was high. COX-2 expression in PCA cells could be suppressed by means of specific phosphatidyl inositol-3 kinase (PI3K), protein kinase C (PKC), or inhibitor of extracellular signal-related kinase (ERK/MAPK) inhibitors. CONCLUSIONS AND CLINICAL IMPORTANCE: COX-2 is expressed in canine PCA; however, expression is not related to the presence of inflammatory infiltrates. This conclusion is further supported by the finding that the cytokines TNF-alpha and IL-6 and their involved signaling pathways do not stimulate COX-2 expression in malignant canine prostate cells.

Identification of candidate genes involved in gonocyte development.

Several germ cell tumors that occur in adult men likely originate from gonocytes that are impaired in their development. To select candidate genes that are involved in the normal development of the gonocytes, we constructed a complementary DNA (cDNA) library from rat gonocytes as well as from single, paired, and aligned A spermatogonia (A(s), A(pr), and A(all) spermatogonia, respectively), their direct descendants. Five hundred gonocyte clones were differentially screened using both libraries. Successive verification by dot blot assays yielded 7 clones that were consistently highly abundant only in the gonocyte cDNA library. Also, in situ hybridization of these clones confirmed their differential expression. They encoded for, respectively, succinate dehydrogenase, ribosomal protein S15a, and the 65-kilodalton scaffolding subunit (a isoform) of protein phosphatase 2A. No clues regarding the nature of the remaining four clones could be found. Hence, using differential screening on both constructed cDNA libraries, we were able to select several genes that are interesting candidates for studying the molecular mechanisms of normal gonocyte development.