Molecular markers allow sibling species identification in red wood ants (Formica rufa group)

Red wood ants (Formica rufa group) constitute a group of species that are considered to be among the most promising bioindicators in forest ecosystems. However, because of their morphological similarity and intraspecific variability, morphological species identification can be difficult. Considerable expertise is necessary to discriminate between the sibling species F. lugubris and F. paralugubris, two species that often live in sympatry in the same Alpine forests. New taxonomic tools providing rapid and reliable species identification are needed. We present a simple and reliable molecular technique based on mtDNA (COI gene) and a restriction enzyme for discriminating between F. lugubris and F. paralugubris. We confirm the validity of this method with a Bayesian analysis based on microsatellites. This new molecular tool represents a clear breakthrough for discriminating between F. lugubris and F. paralugubris and is likely to be helpful in large-scale biomonitoring.

Comparative genomics of epidemic versus sporadic Staphylococcus aureus strains does not reveal molecular markers for epidemicity.

Staphylococcus aureus, especially when it is methicillin resistant, has been recognised as a major cause of nosocomial and community-acquired infections. It has also been shown that certain strains were able to cause clonal epidemics whereas others showed a more incidental occurrence. On the basis of this behavioural distinction, a genetic feature underlying this difference in epidemicity can be assumed. Understanding the difference will not only contribute to the development of markers for the identification of epidemic strains but will also shed light on the evolution of clones. Genomes of strains from two independent collections (n=18 and n=10 strains) were analysed. Both collections were composed of carefully selected, genetically diverse strains with clinically well-defined epidemic and sporadic behaviour. Comparative genome hybridisation (CGH) was performed using an Agilent array for one collection (up to 11 probes per open reading frame - ORF), and an Affymetrix array for the other (up to 30 probes per ORF). Presence and absence information of probe homologues and ORFs was taken for analysis of molecular variance (AMOVA) at the strain and behaviour levels. Not a single probe showed 100% concordant differences between epidemic and sporadic strains. Moreover, probe differences between groups were always smaller than those within groups. This was also true, when the analysis was focussed on presence versus absence of ORF's or when probe information was transformed into allelic profiles. These findings present strong evidence against the presence or absence of a single common specific genetic factor differentiating epidemic from sporadic S. aureus clones.

A new perspective on the evolutionary history of western European Sorex araneus group revealed by paternal and maternal molecular markers.

The species of the common shrew (Sorex araneus) group are morphologically very similar, but have undergone a spectacular chromosomal evolution. We investigate here the evolutionary history of the Sorex araneus group distributed in western Europe. In particular, we clarify the position of a difficult species, S. granarius, using sex-specific (mtDNA and Y-chromosome) markers. The karyotype of S. granarius is generally considered similar to the common ancestor of the restricted group considered here. The mtDNA data (1.4 kb) confirms the close relationship between S. granarius and S. araneus sensu stricto (hereafter S. araneus s.s.), but the Y-chromosome (3.4 kb) produces a quite different picture: S. granarius is closely related to another species, S. coronatus. Comparison of mtDNA and Y-chromosome phylogenies suggests that the genetic and chromosomal evolution in this group are disconnected processes. The evolutionary history of the south-western European populations of the S. araneus group can only be understood considering secondary contacts between taxa after their divergence, implying genetic exchanges by means of hybridization and/or introgression.

Decades of population genetic research reveal the need for harmonization of molecular markers: the grey wolf Canis lupus as a case study

Following protection measures implemented since the 1970s, large carnivores are currently increasing in number and returning to areas from which they were absent for decades or even centuries. Monitoring programmes for these species rely extensively on non-invasive sampling and genotyping. However, attempts to connect results of such studies at larger spatial or temporal scales often suffer from the incompatibility of genetic markers implemented by researchers in different laboratories. This is particularly critical for long-distance dispersers, revealing the need for harmonized monitoring schemes that would enable the understanding of gene flow and dispersal dynamics. Based on a review of genetic studies on grey wolves Canis lupus from Europe, we provide an overview of the genetic markers currently in use, and identify opportunities and hurdles for studies based on continent-scale datasets. Our results highlight an urgent need for harmonization of methods to enable transnational research based on data that have already been collected, and to allow these data to be linked to material collected in the future. We suggest timely standardization of newly developed genotyping approaches, and propose that action is directed towards the establishment of shared single nucleotide polymorphism panels, next-generation sequencing of microsatellites, a common reference sample collection and an online database for data exchange. Enhanced cooperation among genetic researchers dealing with large carnivores in consortia would facilitate streamlining of methods, their faster and wider adoption, and production of results at the large spatial scales that ultimately matter for the conservation of these charismatic species.

Subfamilial and tribal relationships of Ranunculaceae: evidence from eight molecular markers

The first molecular phylogenies of the flowering plant family Ranunculaceae were published more than twenty years ago, and have led to major changes in the infrafamilial classification. However, the current phylogeny is not yet well supported, and relationships among subfamilies and tribes of Ranunculaceae remain an open question. Eight molecular markers from the three genomes (nuclear, chloroplast and mitochondrial) were selected to investigate these relationships, including new markers for the family (two homologs of the nuclear CYCLOIDEA gene, the chloroplast gene ndhF, and the mitochondrial intron nad4-I1). The combination of multiple markers led to better resolution and higher support of phylogenetic relationships among subfamilies of Ranunculaceae, and among tribes within subfamily Ranunculoideae. Our results challenge the monophyly of Ranunculoideae as currently circumscribed due to the position of tribe Adonideae (Ranunculoideae), sister to Thalictroideae. We suggest that Thalictroideae could be merged with Ranunculoideae in an enlarged single subfamily.

Molecular characterization of chromosome termini of the arbuscular mycorrhizal fungus Glomus intraradices (Glomeromycota).

The minimum chromosome number of Glomus intraradices was assessed through cloning and sequencing of the highly divergent telomere-associated sequences (TAS) and by pulsed field gel electrophoresis (PFGE). The telomere of G. intraradices, as in other filamentous fungi, consists of TTAGGG repeats, this was confirmed using Bal31 nuclease time course reactions. Telomere length was estimated to be roughly 0.9 kb by Southern blots on genomic DNA and a telomere probe. We have identified six classes of cloned chromosomal termini based on the TAS. An unusually high genetic variation was observed within two of the six TAS classes. To further assess the total number of chromosome termini, we used telomere fingerprinting. Surprisingly, all hybridization patterns showed smears, which demonstrate that TAS are remarkably variable in the G. intraradices genome. These analyses predict the presence of at least three chromosomes in G. intraradices while PFGE showed a pattern of four bands ranging from 1.2 to 1.5 Mb. Taken together, our results indicate that there are at least four chromosomes in G. intraradices but there are probably more. The information on TAS and telomeres in the G. intradicies will be essential for making a physical map of the G. intraradices genome and could provide molecular markers for future studies of genetic variation among nuclei in these multigenomic fungi.

Molecular imaging by micro-CT: specific E-selectin imaging.

The primary goal of this study was to design a fluorescent E-selectin-targeted iodine-containing liposome for specific E-selectin imaging with the use of micro-CT. The secondary goal was to correlate the results of micro-CT imaging with other imaging techniques with cellular resolution, i.e., confocal and intravital microscopy. E-selectin-targeted liposomes were tested on endothelial cells in culture and in vivo in HT-29 tumor-bearing mice (n = 12). The liposomes contained iodine (as micro-CT contrast medium) and fluorophore (as optical contrast medium) for confocal and intravital microscopy. Optical imaging methods were used to confirm at the cellular level, the observations made with micro-CT. An ischemia-reperfusion model was used to trigger neovessel formation for intravital imaging. The E-selectin-targeted liposomes were avidly taken up by activated endothelial cells, whereas nontargeted liposomes were not. Direct binding of the E-selectin-targeted liposomes was proved by intravital microscopy, where bright spots clearly appeared on the activated vessels. Micro-CT imaging also demonstrated accumulation of the targeted lipsomes into subcutaneous tumor by an increase of 32 +/- 8 HU. Hence, internalization by activated endothelial cells was rapid and mediated by E-selectin. We conclude that micro-CT associated with specific molecular contrast agent is able to detect specific molecular markers on activated vessel walls in vivo.

Separating the contribution of glucocorticoids and wakefulness to the molecular and electrophysiological correlates of sleep homeostasis.

Study Objectives: The sleep-deprivation-induced changes in delta power, an electroencephalographical correlate of sleep need, and brain transcriptome profiles have importantly contributed to current hypotheses on sleep function. Because sleep deprivation also induces stress, we here determined the contribution of the corticosterone component of the stress response to the electrophysiological and molecular markers of sleep need in mice. Design: N/A Settings: Mouse sleep facility. Participants: C57BL/6J, AKR/J, DBA/2J mice. Interventions: Sleep deprivation, adrenalectomy (ADX). Measurements and Results: Sleep deprivation elevated corticosterone levels in 3 inbred strains, but this increase was larger in DBA/2J mice; i.e., the strain for which the rebound in delta power after sleep deprivation failed to reach significance. Elimination of the sleep-deprivation-associated corticosterone surge through ADX in DBA/2J mice did not, however, rescue the delta power rebound but did greatly reduce the number of transcripts affected by sleep deprivation. Genes no longer affected by sleep deprivation cover pathways previously implicated in sleep homeostasis, such as lipid, cholesterol (e.g., Ldlr, Hmgcs1, Dhcr7, -24, Fkbp5), energy and carbohydrate metabolism (e.g., Eno3, G6pc3, Mpdu1, Ugdh, Man1b1), protein biosynthesis (e.g., Sgk1, Alad, Fads3, Eif2c2, -3, Mat2a), and some circadian genes (Per1, -3), whereas others, such as Homer1a, remained unchanged. Moreover, several microRNAs were affected both by sleep deprivation and ADX. Conclusions: Our findings indicate that corticosterone contributes to the sleep-deprivation-induced changes in brain transcriptome that have been attributed to wakefulness per se. The study identified 78 transcripts that respond to sleep loss independent of corticosterone and time of day, among which genes involved in neuroprotection prominently feature, pointing to a molecular pathway directly relevant for sleep function.

Clinical implications of molecular neuropathology and biomarkers for malignant glioma.

Malignant gliomas are currently diagnosed based on morphological criteria and graded according to the World Health Organization classification of primary brain tumors. This algorithm of diagnosis and classification provides clinicians with an estimated prognosis of the natural course of the disease. It does not reflect the expected response to specific treatments beyond surgery (eg, radiotherapy or alkylating chemotherapy). Clinical experience has revealed that gliomas sharing similar histomorphological criteria might indeed have different clinical courses and exhibit highly heterogenous responses to treatments. This was very impressively demonstrated first for oligodendrogliomas. The presence or lack of combined deletions of the chromosomal segments 1p/19q was associated with different benefit from radiotherapy and chemotherapy. We review current molecular markers for malignant gliomas and discuss their current and future impact on clinical neuro-oncology.

Molecular and physiological characterisation of lonotropic receptors in the sacculus in drosophila

Chemosensation is the detection of chemical signals in the environment that enable an animal to make informed decisions about food choice, mate preference or predator detection. Dissecting the molecular and neural mechanisms by which animals detect chemical cues is an important goal towards understanding how they interact with the environment. An attractive system to dissect the mechanisms of chemosensation is the olfactory system. One of the most-investigated olfactory systems is that of Drosophila melanogaster, a model organism that is amenable to a powerful combination of genetic and physiological analyses. Embedded within the antennal olfactory organ of Drosophila is an unusual sensory structure called the sacculus. The sacculus is comprised of three distinct chambers, each lined with several sensilla housing two to three neurons. Previous morphological, anatomical and surgical studies of sacculus neurons have implicated sacculus neurons in chemosensation, hygrosensation and/or thermosensation. While a subset of sacculus neurons have been physiologically characterised as temperature sensors, the role of this organ has remained largely mysterious, due to its inaccessibility to peripheral electrophysiological analysis. Recently a new family of olfactory receptors, the lonotropic Receptors (IRs), was identified. Five IRs are expressed in sacculus neurons providing the first selective molecular markers for these cells. In this thesis I describe the molecular, physiological and anatomical characterisation of these neurons. Genetic labelling of specific populations of sacculus neurons with anatomical (CD8:GFP) reporters has identified neurons in sacculus chambers I and II express IR40a+IR93a together with their co- receptor IR25a, while neurons in chamber III express IR64a with its co-receptor IR8a. Both these sets of neurons project to two distinct glomeruli in the antennal lobe; IR40a neurons project to the column and arm, IR64a neurons project to DC4 and DP1m. Through a live optical imaging screen I showed that these neurons are indeed olfactory and IR64a neurons recognise acidic ligands, while IR40a neurons recognise amine ligands. IR40a and IR64a neurons are in fact composed of anatomically and physiologically distinct subpopulations, strongly implying the existence of other factors that define their functional properties. My thesis identifies the sacculus as a specialised olfactory organ capable of detecting acids and bases, which are of widespread importance to insects. The data from my thesis along with data from other labs show the sacculus is composed of different populations of olfactory sensory neurons and thermosensory neurons. Comparative genomic analysis of sacculus IRs across insects reveals them to be among the most conserved of this receptor repertoire, suggesting that the sacculus represents an evolutionarily ancient insect olfactory acid-base sensor. - La détection des produits chimiques se trouvant dans l'environnement (perception chimiosensorielle) permet à un animal de choisir sa nourriture, son partenaire ou encore d'identifier ses prédateurs. Décortiquer les mécanismes moléculaires et neuronaux grâce auxquels les animaux détectent ces signaux chimiques permet de comprendre comment ces animaux interagissent avec leur environnement. Un système intéressant pour décortiquer ces mécanismes de perception chimiosensorielle est le système olfactif, de la drosophile (Drosophila melanogaster), aussi appelée mouche du vinaigre. C'est un animal modèle très utile grâce à la combinaison d'outils génétiques puissants et d'analyses physiologiques facilement réalisables. Dans l'antenne de la drosophile, qui est l'organe olfactif principal de cet animal, se trouve une structure appelée sacculus. Celui-ci est composé de trois chambres distinctes, chacune comprenant plusieurs sensilles à l'intérieur desquelles se trouvent deux à trois neurones. De précédentes études morphologiques et anatomiques des ces neurones ont déterminé qu'ils sont impliqués dans la perception des odeurs, de l'humidité et de la température. Malgré ceci, la fonction principale de cet organe reste largement inconnue, principalement car il est inaccessible aux analyses électrophysiologiques. Récemment, une nouvelle famille de soixante-six récepteurs olfactifs, nommés Récepteurs lonotropiques (IRs), a été découverte chez la drosophile. Cinq IRs sont exprimés dans les neurones du sacculus. Pour la première fois, une sélection de marqueurs moléculaires est disponible pour l'étude de ces cellules. Dans cette thèse, les caractéristiques moléculaires, physiologiques et anatomiques des neurones du sacculus sont décrites. Ces populations de neurones situés dans le sacculus ont été marquées avec des gènes rapporteurs (CD8:GFP). Ceci a montré que les récepteurs IR40a et IR93a sont exprimés ensemble avec le co-récepteur IR25a dans les chambres I et II, tandis que les neurones de la chambre III expriment IR64a avec son co-récepteur IR8a. Ces deux groupes de neurones projettent vers deux glomérules distincts du lobe antennaire : les neurones IR40a projettent vers la column et le arm, alors que les neurones IR64a projettent vers DC4 et DP1m. Un screen d'imagerie optique a démontré que ces neurones sont en effet des neurones olfactifs, et que les neurones IR64a reconnaissent des ligands acides, tandis que les neurones IR40a reconnaissent des ligands aminés. De plus, les neurones IR40a et IR64a sont séparés en sous-populations distinctes anatomiquement et physiologiquement, et d'autres facteurs permettant de définir leurs propriétés fonctionnelles sont probablement impliqués. Cette thèse identifie ainsi le sacculus comme un organe olfactif spécialisé capable de détecter des acides et amines, lesquels sont très importants pour les insectes. Toutes les données collectées durant cette thèse, combinées aux données d'autres laboratoires, montrent que le sacculus est composé de différentes populations de neurones olfactifs et thermosenseurs. Ces IRs sont très conservés parmi les insectes, suggérant que le sacculus représente révolution d'un ancien détecteur olfactif d'acides et de bases chez l'insecte. - Tous les animaux sont capables de percevoir les signaux chimiques dans leur environnement, comme les odeurs ou le goût, via différents organes. L'odorat est le sens qui permet de percevoir les odeurs, et il est implique des neurones olfactifs qui se trouvent dans le nez des mammifères ou les antennes des insectes. La capacité d'un neurone olfactif à détecter une molécule odorante dépend des types de récepteurs olfactifs qu'il exprime. Il existe deux grandes familles de récepteurs qui perçoivent les odeurs : les Récepteurs Olfactifs, ORs, et Récepteurs lonotropiques IRs, qui détectent différents types d'odeurs avec différents mécanismes. Lorsqu'un récepteur reconnaît une molécule odorante, il convertit ce signal en un signal électrique qui est ensuite transmis au centre olfactif dans le cerveau. La drosophile (Drosophila melanogaster), aussi appelée mouche du vinaigre, est utilisée comme animal modèle pour étudier l'odorat, parce que son génome entier a été séquencé et que ses gènes sont facilement manipulables. De plus, l'anatomie du système olfactif de la mouche est similaire à celui des mammifères, malgré qu'il possède moins de neurones, ce qui le rend moins complexe. Ma thèse a pour objectif d'étudier les Récepteurs lonotropiques dans un organe spécifique, appelé le sacculus, situé dans les antennes. Les neurones du sacculus exprimant des IRs envoient leurs projections au centre olfactif du cerveau, suggérant que ces neurones perçoivent les odeurs. Une technique d'imagerie optique a été utilisée sur le cerveau de mouches vivantes afin de mesurer la réponse des neurones du le sacculus à différentes odeurs. J'ai démontré que ces récepteurs détectent des acides et des amines, qui sont très importants pour les insectes. Par exemple, les acides se retrouvent dans les fruits mûrs sur lesquels les mouches vont se nourrir, s'accoupler et poser leurs oeufs, et les amines sont souvent produites par des bactéries pouvant être nuisible pour la mouche. La principale découverte de ma thèse est donc l'identification du sacculus comme un organe capable de détecter deux des principales odeurs importantes pour la mouche. Ces récepteurs sont aussi présents dans d'autres insectes où ils jouent peut-être des rôles différents. Les acides et les amines se retrouvent aussi dans les excrétions (comme la sueur ou l'urine) de beaucoup de mammifères, qui pourraient potentiellement être dangereux pour la mouche, mais qui attirent les moustiques se nourrissant de leur sang.

Molecular evolutionary patterns in the Glomeromycota;

Abstract Arbuscular mycorrhizal fungi (AMF) form symbiosis with roots of approximately 80% of known land plants. These fungi play a key role in the ecology and adaptation of plants to various ecosystems.by increasing the plant resources for various nutrients. Despite their important ecological role, we still have poor understanding of their genetic structure and their molecular evolution. The work presented in this thesis aims to isolate and analyse AMF genes with various molecular techniques, in order to obtain new insights about their genetics, phylogeny and molecular evolution. Some AMF genes were shown through phylogenetic analyses to be more related with plants or mycoparasites than with other fungal organisms. These results led to the prediction that lateral gene transfers (LGT) occurred between AMF and plants during their long-term co-évolution. By phylogenetic and molecular analyses, in the chapter 2 I demonstrate that the hypothesis of LGT is most likely a consequence of analyses carried out on contaminant non AMF-DNA. In addition, various features characteristic of AMF genes have been determined, allowing researchers to scan their own sequence databases for potential non-AMF contaminants. Phylogenetic relationships of AMF with other fungi has been mostly analysed using molecular markers of ribosomal origin. In chapter 2 I successfully isolated gene encoding α- and ß-tubulins from several AMF genera. Consequently, phylogenetic analyses showed that AMF possess an unexpected relationship with ancestral aquatic fungi (chytrids). These results are consistent with the prediction stating that AMF may have played an important role in the colonisation of land by green plants through the establishment of a symbiosis and after the divergence of AMF from aquatic ancestors. In Chapter 4 I tried to isolate the entire AMF gene family encoding P-Type II ATPases, in order to determine their molecular evolution with the fungal kingdom. These genes were further analysed to detect the level of sequence polymorphism that is present within an AMF population. The results obtained show that mutational events previously thought as occurring only among divergent evolutionary lineages (gene duplications, indel mutations in coding regions) can occur within a single population of AMF. These results have far reaching consequences for our understanding of the genetics and ecology of AMF. Résumé Les champignons endomycorrhiziens arbusculaires (CEA) forment une symbiose racinaire avec environ 80% des plantes vasculaires connues. Ces champignons possèdent un rôle important dans l'écologie et l'adaptation des plantes au sein de différents écosystèmes en .augmentant leurs ressources en nutriments. Le travail présenté dans cette thèse se propose d'isoler et d'analyser certains gènes de CEA avec différentes techniques moléculaires à fin d'obtenir de pÌus amples informations concernant l'évolution moléculaire, la phylogénie et leur diversité génétique à diverses échelles taxonomiques. Certaines analyses phylogénétiques des CEA ont conduit à l'hypothèse que des transferts horizontaux de gènes (THG) ont pu avoir lieu durant leur longue co-évolution avec les plantes vasculaires. Dans le chapitre 2 de cette thèse nous démontrons par analyses moléculaire et phylogénétique que l'hypothèse de THG est une conséquence de contaminations à partir d'ADN de plante ou d'autres micro-organismes. De plus, de nombreuses caractéristiques moléculaires de CEA ont pu être déterminées, permettant la mise en place d'un plan à suivre lors de l'analyse de gènes de CEA dans les études futures. Les relations évolutives des. CEA avec d'autres champignons ont été analysées majoritairement à l'aide de marqueurs moléculaires d'origine ribosomiale. Dans les chapitres 2 et 3 j'ai isolé des gènes codant pour l'a- et la ß-tubuline chez différents genres, de CEA. Les analyses phylogénétiques ont démontré une parenté entre les CEA et des champignons aquatiques ancestraux (chytrides). Ces résultats sont en accord avec l'hypothèse selon laquelle les CEA ont probablement joué un rôle primordial dans l'établissement des plantes sur terre à travers une symbiose et suite à leur évolution à partir d'ancêtres vivant dans des milieux aquatiques: Dans le chapitre 4 j'ai isolé une entière famille de gènes chez les CEA codant des ATPases de la membrane plasmique, et étudié leur évolution moléculaire dans le règne des champignons. Ces mêmes gènes ont été analysés ultérieurement à fin de déterminer le degré de polymorphisme de séquence qui peut être présent au sein d'une population de CEA. Les résultats obtenus montrent que des évènements mutationnels considérés comme apparaissant exclusivement dans des lignées évolutives très divergentes (duplication de gènes, insertions/délétions dans des régions transcrites du génome) ont lieu sein d'une même population de CEA. Cette découverte a un impact important sur nos connaissances concernant la génétique des populations des CEA et leur écologie.

Clinical utility of g-cimp and idh1 status as dual prognostic markers in glioblastoma

The glioma CpG island methylator phenotype (G-CIMP) has been shown to be highly correlated with prognosis andwas noted to be highly concordant with IDH1mutation in malignant glioma in the limited number of samples analyzed. To better understand the relationship of G-CIMP with IDH1 mutation status and patient outcome, we examined G-CIMP status in detail in a larger retrospective series of glioblastomas as well as tumor samples from the RTOG 0525 clinical trial. Sampleswere tested for 6 CIMPmarkers andwere correlated with patient outcomes. In the retrospective tumor set (n ¼ 301),we found 3 distinct survival groups based on the number of CIMP markers: 0-1 (CIMP-negative), 2-4 (CIMP-intermediate), and 5 or greater (CIMP-positive) with median survivals 13.8, 20.1, and 90.6 months, respectively. This finding was validated in the RTOG 0525 samples (median survivals 15.0, 20.3, and 37.0 months). Among 787 cases with both IDH and CIMP data, 617 were CIMP-negative, 136 were CIMP-intermediate, and 34 were CIMP-positive. Seven hundred forty-four were wild type for IDH1 mutation, and 43 were mutant. CIMP and IDH status were positively correlated but outliers were found. Among the 610 CIMP-negative tumors, there were 7 IDH-mutant tumors, which showed no difference in outcome. Similarly, among the 34 CIMP-positive tumors, there were 21 IDH-mutant cases, which also showed no difference in outcome. However, among the CIMP-intermediate cases, there were 15 IDH-mutant cases with significantly (p ¼ 0.0003) improved outcome (medians not reached vs. 18.5 months, 2 year survival 87% vs. 32%). Multivariate analysis showed that both IDH1 mutation status and CIMP status were independent predictors of outcome. These findings suggest the clinical utility of refining the CIMP status into negative, intermediate, and positive groups and the finding that both IDH1 and CIMPstatus are important molecular markers in GBM.

Integrated Analysis of Molecular and Clinical Prognostic Factors in Stage II/III Colon Cancer.

Background The prognostic potential of individual clinical and molecular parameters in stage II/III colon cancer has been investigated, but a thorough multivariable assessment of their relative impact is missing. Methods Tumors from patients (N = 1404) in the PETACC3 adjuvant chemotherapy trial were examined for BRAF and KRAS mutations, microsatellite instability (MSI), chromosome 18q loss of heterozygosity (18qLOH), and SMAD4 expression. Their importance in predicting relapse-free survival (RFS) and overall survival (OS) was assessed by Kaplan-Meier analyses, Cox regression models, and recursive partitioning trees. All statistical tests were two-sided. Results MSI-high status and SMAD4 focal loss of expression were identified as independent prognostic factors with better RFS (hazard ratio [HR] of recurrence = 0.54, 95% CI = 0.37 to 0.81, P = .003) and OS (HR of death = 0.43, 95% CI = 0.27 to 0.70, P = .001) for MSI-high status and worse RFS (HR = 1.47, 95% CI = 1.19 to 1.81, P < .001) and OS (HR = 1.58, 95% CI = 1.23 to 2.01, P < .001) for SMAD4 loss. 18qLOH did not have any prognostic value in RFS or OS. Recursive partitioning identified refinements of TNM into new clinically interesting prognostic subgroups. Notably, T3N1 tumors with MSI-high status and retained SMAD4 expression had outcomes similar to stage II disease. Conclusions Concomitant assessment of molecular and clinical markers in multivariable analysis is essential to confirm or refute their independent prognostic value. Including molecular markers with independent prognostic value might allow more accurate prediction of prognosis than TNM staging alone.