The local Dirichlet process

As a generalization of the Dirichlet process (DP) to allow predictor dependence, we propose a local Dirichlet process (lDP). The lDP provides a prior distribution for a collection of random probability measures indexed by predictors. This is accomplished by assigning stick-breaking weights and atoms to random locations in a predictor space. The probability measure at a given predictor value is then formulated using the weights and atoms located in a neighborhood about that predictor value. This construction results in a marginal DP prior for the random measure at any specific predictor value. Dependence is induced through local sharing of random components. Theoretical properties are considered and a blocked Gibbs sampler is proposed for posterior computation in lDP mixture models. The methods are illustrated using simulated examples and an epidemiologic application.

Serum soluble E-cadherin is a potential prognostic marker in esophageal squamous cell carcinoma

E-cadherin is a well-documented tumor suppressor with downregulated expression in many cancer types. Upon proteolytic cleavage, a soluble form of 80-kDa degradation fragment, known as soluble E-cadherin (s-Ecad), is present in circulation; its level in sera of cancer patients is significantly associated with metastasis, recurrence, and prognosis in some malignancies. The present study investigated the association of s-Ecad with clinicopathological characteristics of patients with esophageal squamous cell carcinoma (ESCC) and its prognostic significance. A cohort of 97 patients who underwent surgery alone (n= 56) or neoadjuvant chemoradiation therapy and surgery (CRT) (n= 41) was recruited for this study. Serum samples were collected at operation (surgery group) and pre- and post-CRT treatment (CRT group) for measurement of s-Ecad protein by enzyme linked immunosorbent assay. Serum s-Ecad levels were correlated with clinicopathological parameters as well as survival. Univariate analysis showed no significant relationship between serum s-Ecad level and clinicopathological parameters for all sets of samples. Survival analysis showed that in patients who had surgical resection only, those with s-Ecad levels equal to or below the median value survived significantly longer than those with levels above the median (median survival 25.6 vs. 14.1 months, P= 0.012). Multivariate analysis showed that pathological N stage, M stage, R category, and serum s-Ecad level were significant independent prognostic factors for ESCC patients who underwent surgery only. The hazard ratio for s-Ecad was 1.104 (95% CI: 1.026-1.187) and P= 0.008. Serum s-Ecad was detected in ESCC patients and its potential as an independent prognostic marker requires further investigation.

Nonparametric Bayes Conditional Distribution Modeling With Variable Selection

This article considers a methodology for flexibly characterizing the relationship between a response and multiple predictors. Goals are (1) to estimate the conditional response distribution addressing the distributional changes across the predictor space, and (2) to identify important predictors for the response distribution change both within local regions and globally. We first introduce the probit stick-breaking process (PSBP) as a prior for an uncountable collection of predictor-dependent random distributions and propose a PSBP mixture (PSBPM) of normal regressions for modeling the conditional distributions. A global variable selection structure is incorporated to discard unimportant predictors, while allowing estimation of posterior inclusion probabilities. Local variable selection is conducted relying on the conditional distribution estimates at different predictor points. An efficient stochastic search sampling algorithm is proposed for posterior computation. The methods are illustrated through simulation and applied to an epidemiologic study.

a5b1 integrin mediates survival of breast cancer cells in 3-dimensional culture that can be specifically targeted for therapy

Abstract #4126

Background: β1 integrin signaling has been implicated in breast cancer progression and has been shown to facilitate resistance to radiation that could be abrogated using β1 integrin inhibitory antibody, AIIB2. In the present study, we investigated whether a specific alpha heterodimer of β1 integrin was preferentially mediating a pro-survival signal in human breast cancer cells, and if it could specifically be targeted for therapy.
 Materials and Methods: We used three-dimensional (3D) laminin-rich extracellular matrix (lrECM) culture to compare the relative levels of alpha heterodimer expression in malignant HMT3522-T4-2 and nonmalignant cell line HMT3522-S-1. Total cell lysates were probed by western blot to detect several possible alpha integrins. In addition, to investigate whether α5β1 mediated an important pro-survival signal, we treated 3D lrECM cultures with a small peptide inhibitor of α5β1 integrin and fibronectin interaction, Ac-PHSCN-NH2 (ATN-161) and measured apoptotic response using TUNEL (terminal deoxynucleotidyl transferase (TdT)-mediated nick end labeling) assay.
 Results: We found that expression of α5β1 was upregulated in HMT3522-T4-2 malignant breast cell line compared to nonmalignant cell line HMT3522-S-1. Inhibition using ATN-161 resulted in a significant dose-dependent increase in TUNEL positive cells.
 Discussion: We have previously shown that inhibitory antibody to β1 integrin, AIIB2, leads to selective apoptosis and decreased proliferation in human breast cancer cells in 3DlrECM culture and in vivo. In addition, combining β1 integrin inhibition with IR allowed for reduction of IR dose necessary to achieve growth inhibition in vivo. Here, we show that in HMT3522-T4-2 breast cancer cell line, inhibition of α5 integrin and fibronectin interaction using ATN-161 significantly enhanced apoptosis in 3D lrECM. These data indicate that fibronectin and its receptor, α5β1 are important for survival signaling in breast cancer and may be an important target to enhance radiotherapy efficacy.

Citation Information: Cancer Res 2009;69(2 Suppl):Abstract nr 4126.

Utilization of allopurinol in the Australian community

BACKGROUND

International data suggest that suboptimal use of allopurinol is common. Allopurinol dose should be lower in renal impairment, but higher when gout is not controlled. The aim of the study was to examine trends in the usage of allopurinol in the Australian community.

METHODS

Community dispensing data on the urate-lowering drugs allopurinol and probenecid were obtained from databases kept by Medicare Australia and the Drug Utilization Sub-Committee, for January 1992 to December 2005.

RESULTS

Allopurinol comprised 98.4% of all prescriptions for urate-lowering drugs dispensed during 2005. Most prescriptions were for allopurinol 300 mg, but there was a steady shift towards use of allopurinol 100 mg in all states and territories over the period of the study. There were marked variations in prescribing rates across the country. New South Wales had the highest rate of subsidized prescribing for allopurinol 300 mg (39.3 per 1000 population). Tasmania had the highest rate for allopurinol 100 mg (14.3 per 1000 population), which coincided with an educational programme to decrease allopurinol dose in patients with renal impairment. Prescribing rates in the Northern Territory were substantially lower than all other regions, at 10.8 and 3.3 prescriptions per 1000 population for allopurinol 300 and 100 mg, respectively.

CONCLUSION

The increased uptake of allopurinol 100 mg suggests greater adherence to dosing guidelines and that there is value in educational programmes to optimize drug usage. Variability in utilization rates across regions indicates the need for research on factors responsible. Precise understanding of dosing trends requires access to deidentified, individual dosing data.

Interference between two indistinguishable electrons from independent sources

Very much like the ubiquitous quantum interference of a single particle with itself, quantum interference of two independent, but indistinguishable, particles is also possible. For a single particle, the interference is between the amplitudes of the particle's wavefunctions, whereas the interference between two particles is a direct result of quantum exchange statistics. Such interference is observed only in the joint probability of finding the particles in two separated detectors, after they were injected from two spatially separated and independent sources. Experimental realizations of two-particle interferometers have been proposed; in these proposals it was shown that such correlations are a direct signature of quantum entanglement between the spatial degrees of freedom of the two particles ('orbital entanglement'), even though they do not interact with each other. In optics, experiments using indistinguishable pairs of photons encountered difficulties in generating pairs of independent photons and synchronizing their arrival times; thus they have concentrated on detecting bunching of photons (bosons) by coincidence measurements. Similar experiments with electrons are rather scarce. Cross-correlation measurements between partitioned currents, emanating from one source, yielded similar information to that obtained from auto-correlation (shot noise) measurements. The proposal of ref. 3 is an electronic analogue to the historical Hanbury Brown and Twiss experiment with classical light. It is based on the electronic Mach-Zehnder interferometer that uses edge channels in the quantum Hall effect regime. Here we implement such an interferometer. We partitioned two independent and mutually incoherent electron beams into two trajectories, so that the combined four trajectories enclosed an Aharonov-Bohm flux. Although individual currents and their fluctuations (shot noise measured by auto-correlation) were found to be independent of the Aharonov-Bohm flux, the cross-correlation between current fluctuations at two opposite points across the device exhibited strong Aharonov-Bohm oscillations, suggesting orbital entanglement between the two electron beams.

Factors affecting decreasing radiation dose for mammography in North Carolina after 2002: an analysis of Food and Drug Administration annual surveys

RATIONALE AND OBJECTIVES

We sought to determine which factors affected the decrease in average glandular dose recorded at the annual U.S. Food and Drug Administration Mammography Quality Standards Act inspections of mammography equipment in North Carolina from 2002 through 2005.

MATERIALS AND METHODS

Average glandular dose, half-value layer, kVp, equipment age, processing speed, and system speed for every mammography unit in the state were collected by Food and Drug Administration-trained state inspectors. A mixed-effect model was used to evaluate the changes of glandular dose over time and to identify the factors associated with these changes.

RESULTS

There was a statistically significant decrease in average glandular dose in North Carolina since 2002. Factors that were statistically significantly associated with this effect were changes in kVp, equipment age, processing speed, and system speed.

CONCLUSION

Average glandular dose for mammography has decreased in North Carolina after 2002. This change is probably at least partially due to the cumulative effect of direct intervention by mammography consultants and statewide educational seminars on mammography quality control.

Effects of a combined alendronate and calcitriol agent (Maxmarvil) on bone metabolism in Korean postmenopausal women: a multicenter, double-blind, randomized, placebo-controlled study

INTRODUCTION

A randomized, double-blind, prospective, 24-week clinical trial was performed to evaluate the effects of a combinative agent, Maxmarvil, of calcitriol (0.5 mug) and alendronate (5 mg) on bone metabolism in postmenopausal women.

METHODS

A total of 217 postmenopausal women with osteoporosis were enrolled; 199 patients were randomly assigned to one of two treatment groups (Maxmarvil group or alfacalcidol group). None of the patients were vitamin-D-deficient, as assessed by serum 25-hydroxyvitamin D (25(OH)D), nor had they received any drugs affecting bone metabolism before enrollment. Bone mineral densities (BMD) of L1-L4 and the femur were measured by dual-energy X-ray absorptiometry (DXA) at the initial assessment and after 6 months of treatment. Serum biochemical assays, including serum calcium, 24-h urinary calcium excretion, and bone turnover markers (both bone-specific alkaline phosphatase [bsALP] and urine N-telopeptide [NTx]), were performed at the baseline and after 3 and 6 months of treatment.

RESULTS

In the Maxmarvil group, the BMD of the lumbar spine increased up to 2.42+/-0.5% from the baseline after 6 months (p<0.05). On the other hand, the change in BMD in the alfacalcidol group was 0.28+/-0.5% after 6 months. There was no significant difference in femoral BMD between the two groups. The levels of bsALP and NTx were significantly lower in the Maxmarvil group than in the alfacalcidol group (-22.04+/-3.9% vs. -11.42+/-2.8% [p<0.05] and -25.46+/-5.2% vs. 1.24+/-6.2% [p<0.001], respectively). Interestingly, there was a significantly smaller amount of 24-h urinary calcium in the Maxmarvil group (p<0.05).

CONCLUSIONS

Our study demonstrates that a combination of calcitriol and alendronate is quite effective in preventing bone loss, with the advantage of lesser hypercalciuric effect of calcitriol in the postmenopausal osteoporotic women.

Fluctuations of Bit-Error-Rate with randomly varying birefringence in optical fibers

We study the fluctuations of the bit-error-rate in optical communication systems by taking into account the polarization mode dispersion. We demonstrate analytically and numerically that the probability distribution function of bit-error-rate possesses an extended algebraic tail in the presence of weak birefringent disorder in the fiber. The characteristics of the probability distribution function also depend on the electrical filter temporal width. For weak birefringent disorder, our numerical simulations show that the probability distribution function contains a longer algebraic regime if the smaller size of electrical filter temporal width is used.

Interaction of solitons through radiation in optical fibers with randomly varying birefringence

Propagation of solitons in optical fibers is studied taking into account the polarization mode dispersion (PMD) effect. We show that the soliton interaction caused by the radiation emitted by solitons due to the PMD disorder leads to soliton jitter, and we find its statistical properties. The theoretical predictions are justified by direct numerical simulations.