The Effect of Time on Root Coverage Outcomes: A Network Meta-analysis.

The stability of root coverage outcomes has gained a great deal of interest. However, insufficient evidence is available, mainly due to limited direct comparisons among different techniques and the small sample size among clinical trials. Therefore, the aim of this study was to propose a mixed-models network meta-analysis (NMA) that includes the novelty of assessing time on root coverage outcomes while simultaneously comparing different surgical approaches. A literature search was performed by 2 individual reviewers to identify randomized clinical trials (RCTs) reporting the outcomes of root coverage procedures of at least 2 time points to estimate the slopes of different treatment approaches. The primary outcomes were the changes in slopes for recession depth (REC), keratinized tissue width (KTW), and clinical attachment level. Sixty RCTs with a total of 2,554 gingival recessions (1,864 patients) were included in the NMA. Connective tissue graft (CTG) and enamel matrix derivative (EMD) approaches provided superior initial REC reduction compared to flap advancement alone. However, only CTG-based procedures were effective in maintaining the stability of the gingival margin over time, while EMD, acellular dermal matrix, collagen matrix, and flap alone showed a similar tendency for gingival recession recurrence. Baseline REC and KTW at the earliest postoperative recall were predictors for the stability of the gingival margin. In addition, a geographic center effect on the treatment slopes was observed for REC and KTW. While limitations of the present linear mixed-modeling approach should be considered as it refers to estimation and comparison of time slopes based on an examined while linear framework, the designed NMA showed to be an effective tool for the simultaneous comparison of multiple treatment approaches while taking into account the critical element of time.

Organ-specific molecular classification of primary lung, colon, and ovarian adenocarcinomas using gene expression profiles.

Molecular classification of tumors based on their gene expression profiles promises to significantly refine diagnosis and management of cancer patients. The establishment of organ-specific gene expression patterns represents a crucial first step in the clinical application of the molecular approach. Here, we report on the gene expression profiles of 154 primary adenocarcinomas of the lung, colon, and ovary. Using high-density oligonucleotide arrays with 7129 gene probe sets, comprehensive gene expression profiles of 57 lung, 51 colon, and 46 ovary adenocarcinomas were generated and subjected to principle component analysis and to a cross-validated prediction analysis using nearest neighbor classification. These statistical analyses resulted in the classification of 152 of 154 of the adenocarcinomas in an organ-specific manner and identified genes expressed in a putative tissue-specific manner for each tumor type. Furthermore, two tumors were identified, one in the colon group and another in the ovarian group, that did not conform to their respective organ-specific cohorts. Investigation of these outlier tumors by immunohistochemical profiling revealed the ovarian tumor was consistent with a metastatic adenocarcinoma of colonic origin and the colonic tumor was a pleomorphic mesenchymal tumor, probably a leiomyosarcoma, rather than an epithelial tumor. Our results demonstrate the ability of gene expression profiles to classify tumors and suggest that determination of organ-specific gene expression profiles will play a significant role in a wide variety of clinical settings, including molecular diagnosis and classification.

Finite mixture modeling with mixture outcomes using the EM algorithm.

This paper discusses the analysis of an extended finite mixture model where the latent classes corresponding to the mixture components for one set of observed variables influence a second set of observed variables. The research is motivated by a repeated measurement study using a random coefficient model to assess the influence of latent growth trajectory class membership on the probability of a binary disease outcome. More generally, this model can be seen as a combination of latent class modeling and conventional mixture modeling. The EM algorithm is used for estimation. As an illustration, a random-coefficient growth model for the prediction of alcohol dependence from three latent classes of heavy alcohol use trajectories among young adults is analyzed.