Wiener klinische Wochenschrift

, Volume 125, Issue 19, pp 577–590

Review on novel concepts of columnar lined esophagus

Authors

  • Johannes Lenglinger
    • Manometry Lab & Upper GI Service, Department of SurgeryUniversity Clinic of Surgery, CCC-GET, Medical University of Vienna, Vienna General Hospital
  • Stephanie Fischer See
    • Manometry Lab & Upper GI Service, Department of SurgeryUniversity Clinic of Surgery, CCC-GET, Medical University of Vienna, Vienna General Hospital
  • Lukas Beller
    • Manometry Lab & Upper GI Service, Department of SurgeryUniversity Clinic of Surgery, CCC-GET, Medical University of Vienna, Vienna General Hospital
  • Enrico P. Cosentini
    • Manometry Lab & Upper GI Service, Department of SurgeryUniversity Clinic of Surgery, CCC-GET, Medical University of Vienna, Vienna General Hospital
  • Reza Asari
    • Manometry Lab & Upper GI Service, Department of SurgeryUniversity Clinic of Surgery, CCC-GET, Medical University of Vienna, Vienna General Hospital
  • Fritz Wrba
    • Institute for Clinical PathologyMedical University of Vienna, Vienna General Hospital
    • Manometry Lab & Upper GI Service, Department of SurgeryUniversity Clinic of Surgery, CCC-GET, Medical University of Vienna, Vienna General Hospital
  • Sebastian F. Schoppmann
    • Manometry Lab & Upper GI Service, Department of SurgeryUniversity Clinic of Surgery, CCC-GET, Medical University of Vienna, Vienna General Hospital
Review

DOI: 10.1007/s00508-013-0418-z

Cite this article as:
Lenglinger, J., See, S., Beller, L. et al. Wien Klin Wochenschr (2013) 125: 577. doi:10.1007/s00508-013-0418-z

Summary

Background

Columnar lined esophagus (CLE) is a marker for gastroesophageal reflux and associates with an increased cancer risk among those with Barrett’s esophagus. Recent studies fostered the development of integrated CLE concepts.

Methods

Using PubMed, we conducted a review of studies on novel histopathological concepts of nondysplastic CLE.

Results

Two histopathological concepts—the squamo-oxyntic gap (SOG) and the dilated distal esophagus (DDE), currently model our novel understanding of CLE. As a consequence of reflux, SOG interposes between the squamous lined esophagus and the oxyntic mucosa of the proximal stomach. Thus the SOG describes the histopathology of CLE within the tubular esophagus and the DDE, which is known to develop at the cost of a shortened lower esophageal sphincter and foster increased acid gastric reflux. Histopathological studies of the lower end of the esophagus indicate, that the DDE is reflux damaged, dilated, gastric type folds forming esophagus and cannot be differentiated from proximal stomach by endoscopy. While the endoscopically visible squamocolumnar junction (SCJ) defines the proximal limit of the SOG, the assessment of the distal limit requires the histopathology of measured multilevel biopsies. Within the SOG, CLE types distribute along a distinct zonation with intestinal metaplasia (IM; Barrett’s esophagus) and/or cardiac mucosa (CM) at the SCJ and oxyntocardiac mucosa (OCM) within the distal portion of the SOG. The zonation follows the pH-gradient across the distal esophagus. Diagnosis of SOG and DDE includes endoscopy, histopathology of measured multi-level biopsies from the distal esophagus, function, and radiologic tests. CM and OCM do not require treatment and are surveilled in 5 year intervals, unless they associate with life quality impairing symptoms, which demand medical or surgical therapy. In the presence of an increased cancer risk profile, it is justified to consider radiofrequency ablation (RFA) of IM within clinical studies in order to prevent the progression to dysplasia and cancer. Dysplasia justifies RFA ± endoscopic resection.

Conclusions

SOG and DDE represent novel concepts fusing the morphological and functional aspects of CLE. Future studies should examine the impact of SOG and DDE for monitoring and management of gastroesophageal reflux disease (GERD).

Keywords

Columnar lined esophagusBarrett’s esophagusGastroesophageal reflux disease

Übersicht zu neuen Konzepten des Zylinderepithel-Ösophagus

Zusammenfassung

Hintergrund

Zylinderepithel-Ösophagus (engl. columnar lined esophagus; CLE) zeigt gastroösophagealen Reflux und bedingt bei jenen mit einem Barrett Ösophagus ein erhöhtes Krebsrisiko. Rezente Studien beschreiben ein integriertes morphofunktionales CLE Konzept.

Methodik

Diese PubMed basierte Analyse gibt eine Übersicht zu neuen histopathologischen Konzepten zu CLE ohne Dysplasie.

Ergebnisse

Unsere neue Vorstellung zu CLE wird anhand von zwei neuen histopathologischen Konzepten dargestellt: dem Mukosasegment zwischen Plattenepithel und oxntischer Magenschleimhaut (engl. squamo-oxntic gap; SOG) und dem dilatierten distalen Ösophagus (engl. dilated distal esophagus; DDE). Als Folge des Reflux entsteht das SOG zwischen dem von Plattenepithel ausgekleideten Ösophagus und des von oxyntischer Mukosa ausgekleideten proximalen Magens. SOG beschreibt die Histologie des CLE im tubulären Ösophagus und DDE, welcher auf Kosten des durch den Reflux verkürzten unteren Ösophagussphinkters entsteht und damit vermehrten Rückfluss des sauren Mageninhalts begünstigt. Morphologische Untersuchungen des Ausgangs der Speiseröhre zeigten, dass der DDE Reflux-geschädigter, dilatierter, magenähnliche Falten bildender Ösophagus ist und in der Endoskopie nicht vom proximalen Magen unterschieden werden kann. Während die proximale Grenze des SOG der endoskopisch definierbaren Platten-Zylinderepithelgrenze entspricht, kann die untere Grenze des SOG nur mittels Fusion von Biopsie-Lokalisation und der Histologie von aus diesem Bereich entnommenen Gewebeproben bestimmt werden. Im SOG ordnen sich die CLE Typen entsprechend einer typischen proximalen-distalen Verteilung mit intestinaler Metaplasie (IM, Barrett Ösophagus) ± Kardia Schleimhaut (CM) an der Platten-Zylinderepithelgrenze und Oxyntokardia (OCM) Mukosa im distalen Abschnitt des SOG. Die Ausrichtung folgt dem Reflux-bedingte pH Gradienten entlang des unteren Ösophagus. Die Diagnose von SOG und DDE erfolgt mittels Endoskopie, Histologie von Multi-Level Biopsien aus dem Ausgang der Speiseröhre sowie Funktionstests und Röntgenuntersuchungen. CM und OCM an sich bedürfen keiner Therapie und sollen in 5 Jahren nachuntersucht werden, nur assoziierte Reflux Beschwerden, welche die Lebensqualität beeinträchtigen, sollen medikamentös oder chirurgisch behandelt werden. Bei entsprechendem Krebsrisiko ist es gerechtfertigt, bei IM ohne Dysplasie eine Radiofrequenzablation (RFA) im Rahmen klinischer Studien zu erwägen, um damit die Entstehung von Dysplasie und Karzinom zu verhindern. Dysplasie rechtfertigt eine RFA ± endoskopischer Resektion.

Schlussfolgerungen

SOG und DDE sind neue Konzepte, welche Morphologie und Funktion des Zylinderepithel-Ösophagus integrieren. Die Zukunft wird zeigen, welche Bedeutung diese neuen Konzepte für Diagnose und Therapie der gastroösophagealen Refluxkrankheit haben.

Schlüsselwörter

Zylinderepithel ÖsophagusBarrett ÖsophagusEndoskopieGastroösophageale RefluxkrankheitHistopathologie

Introduction

Gastroesophageal reflux disease (GERD) affects 20–30 % of the population in Europe and North America (annual incidence 0.5 %) [13]. In addition to the impairment of the life quality and productivity due to the symptoms (heartburn, regurgitation, cough, wheezing, and asthma) [48], reflux alters the morphology of the esophagus [910]. As a consequence of reflux induced inflammation the normal squamous lining of the esophagus is replaced by a columnar epithelium, which is termed as columnar lined esophagus (CLE) [1115]. Furthermore, the morphologic changes are paralleled by an incompetence of the antireflux mechanism (lower esophageal sphincter—LES) and an impairment of the transport function of the esophagus [1619]. CLE attracts our attention, because it may become the platform for cancer development among those with Barrett’s esophagus [2022]. Finally, CLE not necessarily develops in conjunction with GERD symptoms and may thus also be present in asymptomatic persons [2325]. This fact is mirrored by the observation that the majority of reflux induced cancers of the esophagus develop without a prior history of GERD symptoms [20, 26]. Therefore, this review aims to summarize our current understanding regarding the pathophysiology of nondysplastic CLE and the implications for diagnosis and treatment.

Methods

We conducted a PubMed research on studies examining novel concepts of CLE. The review mainly focuses on studies published within the last 3 years (2010–2012). Special attention is given to novel concepts of esophageal pathophysiology, including the dilated distal esophagus (DDE) and the squamo-oxyntic gap (SOG). Novel treatments included radiofrequency ablation (HALO®, GI Solutions, Covidien) and the surgical magnetic sphincter augmentation (LINX® System, Torax). Statistics were not applied.

Results

Normal foregut anatomy

The esophagus lies within the mediastinum, including a cervical, thoracic, and abdominal portion and represents a muscle tube for the transport of food from the neck into the stomach [27]. Normally stratified squamous epithelium lines the luminal surface of the esophagus [28]. A muscular upper and lower esophageal sphincter compartmentalize the esophagus from the throat and stomach, respectively [27]. While the upper esophageal sphincter works under conscious control, the lower esophageal sphincter opens only during the passage of food, belching, and vomiting [27]. Otherwise, the LES is closed and thus prevents the reflux of gastric content [27]. The oxyntic mucosa (straight tubular glands extending down to the level of the muscularis mucosae, presence of parietal, chief cells, and absence of mucus cells within the subfoveolar region of the glands) covers the inner surface of the proximal stomach (fundus, corpus) [11, 12, 29, 30]. A mucus cell only mucosa covers the distal portion of the stomach (antrum) [11, 12, 29]. In contrast to the stomach, the esophagus contains submucosal glands and lacks a peritoneal coverage [11, 12, 27, 28, 30]. Therefore, the peritoneal reflection (the phrenicoesophageal ligament) represents the anatomical junction between the esophagus and the stomach [11, 12, 27, 30].

The dilated distal esophagus

Fusion of foregut anatomy, histopathology, and data obtained from esophageal function tests in GERD patients [1619, 31] and normal individuals [30] contributed to our novel understanding of the pathophysiology underlying the formation of CLE (Fig. 1).

https://static-content.springer.com/image/art%3A10.1007%2Fs00508-013-0418-z/MediaObjects/508_2013_418_Fig1_HTML.jpg
Fig. 1

The antegrade endoscopic images and the corresponding schematic drawings cartoon the pathophysiology underlying the development of columnar lined esophagus (CLE) and the dilated distal esophagus (DDE), as described in the text. a Gastric distention-induced shortening of the lower esophageal sphincter (LES; green) causes reflux into the intra-sphincteric portion (panel B). The DDE, this is the dilated CLE portion of the lower end of the esophagus, develops at the cost of the LES (b, c). Over time LES incompetency allows reflux into the supra-sphincteric portion of the esophagus. As the DDE extends towards the level of the diaphragm, it widens the hiatus and forms the hiatal hernia (c). With increasing length of the DDE/CLE the squamocolumnar junction (SCJ) dislocates proximally (yellow circle: level of diaphragmatic impression; yellow asterisk: level of the rise of the endoscopically visiblegastric type fold”; red arrow: SCJ); D diaphragm, E esophagus. LES: lower esophageal sphincter

Recent studies indicate, that reflux occurs, if the antireflux mechanism (LES) within the lower portion of the esophagus fails to function [17, 19]. This happens if a person repeatedly over eats/drinks and thus distends the stomach (i.e., carbonated beverages) [17] (Fig. 1a). As a consequence, gastric dilatations propagate into the lower portion of the esophagus and transiently overstretch and shorten the sphincter apparatus [17] (Fig. 1b). This in turn causes repeated, swallow-independent transient relaxations and shortenings of the abdominal portion of the antireflux mechanism (i.e., transient lower esophageal sphincter relaxations—TLESR) [17]. At this early stage, transient episodes of acid gastric reflux into the LES occur (i.e., intrasphincteric reflux) (Fig. 1b). At some critical point the stretching, relaxation, and shortening of the sphincter becomes permanent (permanent sphincter dilatation) and the outlet of the esophagus stays permanently open and gains the shape of a trumpet like opening (“vuvuzela sign”) [17, 32] (Fig. 1b). Consequently the dilated segment of the lower portion of the esophagus gets permanently exposed to gastroesophageal reflux (Fig. 1b). Furthermore, these events set the stage for reflux to extend into the distal esophagus above the level of the LES (i.e., supra-sphincteric reflux), which is detected 5 cm and ³ 3 cm above the manometric LES during the classical pH-monitoring and the combined impedance pH test, respectively [1619] (Fig. 1c). The incompetence of the LES is paralleled by a decreased resistance to withstand elevations of the intragastric pressure. Thus, in an incompetent sphincter, TLESRs are induced by a significantly lower intra-gastric pressure, when compared to a competent LES [17]. These observations recently motivated the development of a novel antireflux operation [33]. Here the application of a ring composed of magnetized titan pearls is thought to restore the abdominal length of the LES and to impair gastric distention induced unfolding of the lower end of the esophagus, i.e., inhibit TLESR. Thus the magnetic sphincter prevents reflux and GERD symptoms [33].

The above functional alterations, (i.e., the sphincter shortens at the cost of its abdominal portion) (Fig. 2) are paralleled by typical morphologic changes [16, 17, 32]. Reflux inflames the squamous lined mucosa (esophagitis) [28, 29] which in turn causes the mediation of an inflammatory response involving all cells and mediators released from the cells within the esophageal mucosa and the submucosa [34, 35]. In majority of the cases the inflammation does not involve the muscle layers of the esophagus [1113]. Only in severe cases the esophagitis extends into the mediastinum and the diaphragmatic crura [31]. Recent animal studies indicated the involvement of bone marrow derived stem cells in the mediation of esophagitis [36].

https://static-content.springer.com/image/art%3A10.1007%2Fs00508-013-0418-z/MediaObjects/508_2013_418_Fig2_HTML.gif
Fig. 2

Combined impedance (upper violet bolus tracing) and high resolution manometry color plot (lower tracing) of a person with symptoms of gastroesophageal reflux disease (GERD). Swallow (yellow arrow) induced bolus transport (violet tracing) and contraction waves (lower panel) are normal. Note the presence of a shortened esophagogastric junction high pressure zone (HPZ) at the level of the diaphragm with a transient pressure fall upon swallowing. The swallow induced contraction wave runs out at the level of the HPZ, which corresponds to the shortened lower esophageal sphincter (LES). The lack of HPZ distal to the level of the diaphragm (red arrow) indicates the presence of a dilated distal esophagus (DDE) including the abdominal portion of the LES (corresponds to Fig. 1b). White asterisk marks an intra-thoracic high pressure band corresponding to cardiac and/or aortic pulsations. Tracings obtained, using Sandhill technology

The reflux induced neurohumoral flush stresses the esophagus [34, 35], which in turn is suggested to impair the function of the LES [16], to drive the genetic program towards the formation of CLE [21], and to mediate the perception of symptoms (heartburn) [28]. It is not clear if the reflux induced incompetence of the sphincter results from a direct caustic injury of the muscle, the indirect action of the neurohumoral flush, or both [34, 35]. As a consequence of the loss of sphincter competence, the esophagus irreversibly dilates and forms longitudinal folds covered by a columnar lined mucosa (CLE) [32] (Fig. 1c). Now the lower end of the esophagus gains a “gastric type” appearance (i.e., a reservoir composed of columnar mucosa lined longitudinal gastric type folds), called DDE [32] (Fig. 1b, c). Interpretation of the data of recent investigations justifies to suggest that the DDE develops at the cost of the abdominal portion of manometric LES, i.e., the shortening of the abdominal portion of the manometric LES indicates the presence of DDE [17, 32] (Figs. 2 and 3).

https://static-content.springer.com/image/art%3A10.1007%2Fs00508-013-0418-z/MediaObjects/508_2013_418_Fig3_HTML.gif
Fig. 3

Combined impedance (upper violet bolus tracing) and high resolution manometry color plot (lower tracing) of a person with symptoms of gastroesophageal reflux disease (GERD). Swallow (yellow arrow) induced bolus transport (violet tracing) and contraction waves (lower panel) are normal. Note the shortened esophagogastric junction high pressure zone (HPZ) above the level of the diaphragm. The swallow induced contraction wave runs out above the level of the diaphragm (red arrow). This indicates the presence of a dilated distal esophagus (DDE) including the abdominal and most of the thoracic portion of the LES (corresponds to Fig. 1c). Tracings obtained, using Sandhill technology

At this stage the endoscopically visible squamocolumnar junction (SCJ) and the level of the rise of the gastric type folds (i.e. rugal folds; latin “ruga” = fold) coincide with the level of the diaphragmatic impression during endoscopy (Hill grades I, II) [3739] (Fig. 1b). Over time the DDE extends above the level of the diaphragm, i.e., the level of the rise of the endoscopically visible gastric type folds dislocates above the level of the diaphragmatic impressions (Hill valve vanishes and opens during endoscopy; Hill III, IV) [4045] (Fig. 1c). At this point the DDE takes up the entire abdominal and thoracic (supradiaphragmatic) portion of the LES [13, 32, 39, 40, 42] (Fig. 1c). As a consequence manometry fails to assess the lower esophageal sphincter high pressure zone [16, 18, 19, 40, 43] (Fig. 3). Going in line with our novel understanding, the repeated lateral stretch of the DDE during overeating causes a lateral displacement of the diaphragmatic crura and thus enlarges the diameter of the esophageal hiatus within the diaphragm, as a consequence of which the hiatal hernia forms [4045]. The observation, that the hiatal hernia associates with an incompetent, shortened antireflux mechanism suggests, that, like the DDE, the hiatal hernia also develops at the cost of the competency of the manometric LES [17, 4045]. Thus, an endoscopically visible hiatal hernia indicates the presence of DDE and LES incompetence [40, 4246]. At some point the proximal dislocation of the SCJ becomes visible for the endoscope; this is endoscopically visible CLE, interposed between the SCJ and the level of the rise of the gastric type folds [1016, 3739]. Over a period of time the endoscopically visible CLE (CLEv) increases in length and forms islands, tongues, and segments within the tubular esophagus[10, 3739, 47, 48] (Fig. 1c).

The endoscopic “gastric type” appearance of the DDE explains, that it has recently been mistaken for proximal stomach, i.e., the gastric cardia [10, 14, 15, 49, 50]. The length of the DDE cannot be defined by endoscopy. Assessment of the length of the DDE requires the histopathology of measured multi level biopsies obtained from the endoscopic esophagogastric junction, where CLE and oxyntic mucosa define esophageal and gastric location of the biopsies, respectively [1113, 16, 31, 38, 39].The fusion of anatomy, endoscopy, and histopathology revealed that the length of the DDE ranges between < 0.5 cm in children [29] and up to 3 cm in adults [30, 51]. Discrepancy exists whether CLE is an acquired condition. Recent studies demonstrated that biopsies obtained from the SCJ contained the direct transition from squamous lined esophagus to gastric oxyntic mucosa in asymptomatic children [29, 30], indicating the absence of CLE in young individuals. These data support the notion that CLE represents an acquired condition [1113, 52].

The Chandrasoma classification of nondysplastic CLE is the basis for the novel concepts and lists cardiac mucosa (CM; mucus cell only epithelium), oxyntocardiac mucosa (OCM; mixture of mucus cells and parietal cells within the subfoveolar region of the glands), and intestinal metaplasia (IM; columnar epithelium composed by a mixture of goblet cells and mucus cells) [1113, 16, 30, 32, 53] (Fig. 4). In up to 20 % of the cases, biopsies obtained from the SCJ contain a multilayered mixture of squamous and columnar epithelium. According to Glickman et al. this condition is termed “multilayered epithelium (MLE)” [29, 54, 55]. CLE with IM defines nondysplastic Barrett’s esophagus [21]; exception UK and Japan, where CLE per se defines Barrett’s esophagus [56]. IM associates with an increased cancer risk [21, 56]. Dysplasia and cancer are diagnosed according to Riddell and Odze [1557].

https://static-content.springer.com/image/art%3A10.1007%2Fs00508-013-0418-z/MediaObjects/508_2013_418_Fig4_HTML.jpg
Fig. 4

Antegrade endoscopic images and the corresponding schematic drawings of the squamo-oxyntic gap (SOG) (a–c). Panel A cartoons normalcy, i.e., the absence of SOG, as described in the review. The endoscopically visible squamocolumnar junction (SCJ) resides below the level of the diaphragmatic impression. Panel B: As a consequence of the gastric distention-induced reflux the SOG, i.e., the columnar lined esophagus (CLE), interposes between the squamous lined mucosa of the esophagus and the oxyntic mucosa of the proximal stomach, as described in the text. Initially the SOG comprises the dilated distal esophagus (DDE). Panel C: Over time the SOG extends into the tubular esophagus and becomes visible for the endoscope, i.e., the endoscopically visible CLE (CLEv). With increasing length of the DDE/CLE the SCJ dislocates proximally (yellow circle: level of diaphragmatic impression; yellow asterisk: level of the rise of the endoscopically visible “gastric type fold”; red arrow: SCJ; D diaphragm)

The squamo-oxyntic gap (SOG)

According to recent studies, the distance between the most proximal level of the SCJ and the distal limit of the DDE defines SOG [53] (Fig. 5). As a consequence of the reflux the SOG, i.e., CLE, interposes between the squamous lined esophagus and the proximal limit of the proximal stomach [1113, 53]. Going in line with the above suggestions, the SOG develops at the cost of the antireflux mechanism within the distal esophagus (LES and supra-sphincteric portion of the tubular esophagus) [1719, 42, 44]. Consequently, the SOG [53] includes the length of the endoscopically visible CLE (CLEv) within the tubular esophagus [10] and the DDE [32] (Fig. 5c). In contrast to that, the Prague classification only diagnoses the CLEv within the tubular esophagus (islands, circumferential segments, and tongues of columnar mucosa within the tubular esophagus) [10]. However, as a major drawback, the Prague classification [10] repeats the misconception of Norman Barrett [49] and allocates the cardia to the stomach. As a consequence the Prague classification excludes the platform for the development of distal esophageal cancer, i.e., the adenocarcinoma of the dilated distal esophagus [20, 31, 58, 59].

https://static-content.springer.com/image/art%3A10.1007%2Fs00508-013-0418-z/MediaObjects/508_2013_418_Fig5_HTML.jpg
Fig. 5

Histopathology of biopsies obtained from the squamocolumnar junction (SCJ) (a, b) and endoscopically visible columnar lined esophagus (c) in patients with symptoms of gastroesophageal reflux disease (GERD). a Transition from squamous (Squ) to columnar lined esophagus (CLE) with cardiac mucosa (CM) and oxyntocardiac mucosa (OCM). The distribution of the mucosal types follows a distinct proximal to distal zonation with CM proximally and OCM distally, as described in the text. b Direct transition from Squ to OCM, a finding in 10 % of GERD patients. Yellow arrows in panels A and B mark parietal cells in OCM. C: nondysplastic Barrett’s esophagus, white arrow marks a goblet cell, the hallmark of Barrett’s esophagus (H&E stain, magnification a 20x, b, c 50x)

Taken together, the SOG describes the morphologic correlation of the reflux induced “inflammatory foregut disease” of the esophagus and includes the endoscopically visible CLE and the DDE with and without a hiatal hernia [46, 53] (Fig. 5).

From squamous to columnar lined esophagus

We next summarize our current understanding regarding the pathophysiology underlying the development of CLE [60]. Bone morphogenetic protein (BMP) -4 mediates the development of cardiac mucosa [61, 62], which in turn may progress towards IM (CDx2 mediated) or regress towards OCM (mediated by the sonic hedge hoc pathway) [63]. In contrast to CM [21, 56, 58, 59], OCM is not considered to progress to IM and thus does not harbor a cancer risk, unless it transforms to CM (0.07 % annual cancer risk) [56]; i.e., within a given CLE gland the genetic program decides either for parietal or goblet cell [56, 5963].As a consequence, parietal and goblet cells do not co-localize within a given CLE gland [63]. Due to the pH gradient across the distal esophagus, IM always develops at the SCJ [64]. As a consequence, biopsies including the SCJ have the highest yield for the assessment of IM. Due to the pH gradient, the distribution of the mucosal types within the SOG follows a distinct zonation with CM/IM and OCM in the proximal and distal portion, respectively [64] (Fig. 4a). In 10 % and 90 % of the cases the SOG contains OCM or a mixture of OCM/CM ± IM, respectively [30, 3739, 52].

We care about IM because it associates with an increased risk for the development of adenocarcinoma of the esophagus [6570]. Via low- (LGD) and high grade dysplasia (HGD), IM may progress to cancer [2022, 26, 56, 57]. “Indefinite for dysplasia” describes a condition, where the inflammation severely distorts the mucosal architecture and the cellular shape of the epithelial cells [12, 13, 39, 57]. As a consequence the epithelium cannot be accurately defined as either dysplastic or non-dysplastic. Re-biopsy sampling after 2 weeks treatment with proton pump inhibitor is recommended [12, 13, 39].

IM affects 20–30 % of persons with GERD symptoms (3:1 male:female ratio) [13, 21, 52, 53] and up to 25 % of asymptomatic individuals [25]. The prevalence of IM in the normal population ranges between 1.6 % [71], 5.5 % [72] and 16 % [13]. In those with GERD symptoms, the likelihood to detect IM increases with longer segments of endoscopically visible CLE (CLEv). Thus IM is assessed in 10–17 % of the cases in the absence of CLEv and 50, 80, and 100 % in the presence of CLEv of 3 cm, < 5 cm, and ³ 5 cm, respectively [30, 39, 47, 53].

The reported annual cancer risk for IM ranges from 0.12 % [66] to 0.7 % [73] and that of CLE without IM is 0.07 % [56]. IM and cancer is usually detected between 50 and 60 years of age, respectively [74]. Consequently the time from the assessment of IM to the development of the cancer ranges from 5 to 15 years [6570]. The annual cancer risk of LGD ranges between 10 and 15 % [6570, 75].Additional risk factors for cancer development of Barrett’s esophagus (IM) without dysplasia include male gender, esophagitis, progression to LGD/HGD, hiatal hernia > 2.0 cm, CLEv ³ 2.0 cm, history of GERD symptoms ³ 10 years, and administration of proton pump inhibitor therapy (HR 3.5) [69].

During the last 20 years the incidence of esophageal adenocarcinoma significantly increased in western countries (3–5-fold), whereas the numbers for gastric carcinomas decreased [7678]. Currently there seems to be a stabilizing trend in some countries and further increase in other European countries (UK) [7678].

Diagnosis of nonmalignant CLE

Diagnosis establishes the different ways how CLE manifests and associates with impairment of the life quality due to the GERD symptoms and cancer risk [12]. Conceptually, treatment should be offered if CLE harbors an increased cancer risk (IM) and/or associates with GERD symptoms, which impair the life quality (CAVE: increased cancer risk for asymptomatic IM; see below) [12]. In addition to the assessment of the patient history and the symptoms (heartburn, acid regurgitation, dysphagia, wheezing, coughing, throat-, neck and non-cardiac chest pain, and asthma), the diagnostic tests for CLE include: endoscopy and biopsy sampling (esophagogastroduoneoscopy; EGD), esophageal function test (high resolution manometry + intraluminal impedance, impedance pH testing), and radiologic investigations.

Patient history

The study of the patient history aims to elucidate the intensity and frequency of GERD symptoms and how these symptoms impair the life quality and productivity. Within academic studies we recommend the use of life quality assessment tools (i.e., SF-12, SF-36, and gastrointestinal Eypasch score) [5, 6, 79].

Endoscopy

During EGD the physician assesses the presence or absence of endoscopically visible esophagitis, CLE, concomitant morphological abnormalities of the esophagus, (i.e., webs, rings, diverticula, polyps, ulcer, stenosis, varices), and should obtain biopsies from the distal esophagus and the esophagogastric junction [10, 3942, 4548]. Going in line with the data obtained in recent biopsy studies, what is often taken as esophagogastric junction during endoscopy, i.e., the level of the rise of the so called “gastric type folds” in persons with GERD symptoms in 99–100 % of the cases represents the proximal limit of the DDE [29, 30, 32, 39, 52]. Biopsies obtained from this level contain OCM and CM ± IM [2932, 39, 52]. In persons with GERD symptoms, IM (Barrett’s esophagus without dysplasia) is present at a normal appearing esophagogastric junction in 10–17 % of the cases [30, 3739, 52, 53]. The prevalence of IM increases with greater length of endoscopically visible CLE (i.e., 100 % for visible CLE > 5.0 cm) [3037]. However, regarding the oncologic playground the proximal limit of the DDE represents the peak of the iceberg. Recent studies from the Siewert group in Munich demonstrated that the majority of distal esophageal cancers arise within the DDE [80]. Consequently we recommend to extend the biopsy protocol towards the DDE, i.e., the proximal portion of the endoscopically visible gastric type folds [39, 52]. Thus, EGD should follow the Prague classification and catalog the length of segmental, circumferential endoscopically visible CLE (CLEv) and obtain biopsies from a normal appearing junction and from CLEv in 1.0 cm increments [10, 37, 39, 52]. Samples from each biopsy level should be processed as separate specimens. In addition to the Prague classification we recommend to take measured biopsies from the proximal portion of the dilated distal esophagus [39, 52]. Fusion of biopsy site and histopathology (i.e., CLE vs. oxyntic mucosa of the proximal stomach) assesses the length of the DDE [32]. Fusion of the Prague classification (optical and tissue sampling) biopsy sites and the length of the DDE define the entire dimension of the SOG, i.e., length of CLE interposed between the normal squamous epithelium and the proximal stomach as a consequence of the reflux [10, 32, 52, 53]. Most importantly the diagnosis of CLE requires a routinized endoscopist with adequate knowledge of esophageal anatomy, physiology, and pathology and an expert pathologist [14, 15]. This contributes to decrease insecurity and inter-observer disagreement [75]. As a consequence, management of CLE should be centered to institutions with the respective focus, interest and routine in the related field. [14, 22, 60].

Up to 25 % of persons without GERD symptoms harbor IM in their distal esophagus [24, 25]. The large majority of persons with adenocarcinoma of the esophagus present with tumor induced dysphagia and lack a prior history of heartburn and regurgitation [26]. Thus, we only detect 7–10 % of reflux induced cancers via surveillance of Barrett’s esophagus [2426, 66, 81]. The clinical relevance of this finding remains to be questioned (see below).

Approximately 10 % of persons with GERD symptoms harbor CLEv within the proximal esophagus adjacent to the level of the upper esophageal sphincter [82]. CLE at this level is suggested to result from (supine, night time) reflux trapped by the upper sphincter and thus gets exposed to the mucosa of the proximal esophagus [82, 83].In majority of the cases CLEv at this cervical location contains CM and OCM and in less than 5 % of the cases IM or LGD [83]. As a consequence we recommend accurate inspection of the proximal esophagus and biopsy sampling in the presence of CLEv. In rare cases the CLEv in the cervical portion of the esophagus contains gastric oxyntic mucosa (i.e., cervical inlet patch) [84].

During endoscopy the condition where the gastric type folds arise 2 cm and more above the level of the diaphragmatic impression is considered as hiatal hernia [10, 4046]. Based on the above data, the hernia contains DDE (and not stomach) in the majority of persons with CLE [32, 39, 5053]. The retroflexed view towards the esophagogastric junction describes the integrity of the so called Hill valve [39, 41]. Increased Hill grade (impaired function of the valve) correlates with advanced disease, i.e., increased length of endoscopically visible CLE, impaired function of the esophagus, and increased reflux, as assessed by endoscopy, manometry, and impedance pH monitoring, respectively [4345].

Esophageal function tests

CLE develops when the antireflux mechanism and the transport function of the esophagus fail (dysphagia, high volume reflux). Esophageal high resolution manometry color plots the pressure profile of the antireflux mechanism; i.e., the LES or the esophagogastric junction high pressure zone (EGJ HPZ) [8, 18, 19, 44, 45]. The test assesses the length (abdominal, thoracic, entire) of the LES and the mean pressure within the proximal portion of the LES (i.e., at the level of the respiratory inversion point). Increased length of CLE (SOG) associates with decreased function of the performance of the sphincter [16, 17, 19]. In addition, the contour, velocity, and pressure profile of the contraction waves reflects esophageal transport function. However, combination of manometry and intraluminal impedance represents the most accurate tool for the assessment of esophageal transport (and the capacity to clear reflux) [8, 18] (Figs. 2 and 3). Intraluminal impedance measures fluid-transport-induced resistance changes along a catheter placed within the lumen of the esophagus [8, 18, 19, 85]. Consequently, combination of manometry and impedance test is advantageous and enables to differentiate between complete and incomplete bolus transport across the esophagus, irrespective of the pressure events [8, 19, 85] (Figs. 2 and 3). Recent combined tests demonstrated the clinical relevance, as abnormal manometric contraction waves did associate with normal transport, as assessed by intraluminal impedance testing of the esophagus [85]. Manometry should be conducted if antireflux surgery for treatment of symptomatic CLE is considered or other functional disorders of the esophagus are to be ruled out (achalasia, distal esophageal spams, scleroderma) [8, 44, 45, 85].

Impedance pH test (reflux monitoring) assesses both acid and non acid reflux and, if reflux causes the symptoms (heartburn, coughing) which impair the life quality of a patient (Table 1) [86, 87]. Since CLE and gastric mucosa share comparable electrophysiological characteristics (resistance of 200–300 Ω), impedance fails to detect fluid transport (influx, reflux) within segments of CLE [88]. In contrast, impedance enables to discriminate between normal and inflamed squamous epithelium (2000 vs. 500–1,000 Ω for normal esophagus and esophagitis, respectively) [87, 88]. Impedance pH test should be conducted prior to antireflux surgery to define reflux as the cause for the symptoms and serve for treatment monitoring (effect of proton pump inhibitor, antireflux monitoring).

Table 1

Clinical relevance of the histopathology of nonmalignant columnar lined esophagus (CLE)

CLE type

What it tells

Information on symptomatic or asymptomatic reflux?a

CM

History of reflux; 0.07 % annual cancer risk

No

OCM

History of reflux; no cancer risk

No

MLE

History of reflux; no cancer risk

No

IM

History of reflux; 0.12–0.7 % annual cancer riskb

No

LGD

History of reflux; 10–15 % annual cancer riskb

No

HGD

25 % annual cancer riskb

No

CLE without increased cancer risk does not require treatment, unless it is associated with symptom-induced impairment of the life quality; here the therapy is directed against the symptoms (life style, medical, surgery), as described in the text

CM cardiac mucosa, OCM oxyntocardiac mucosa, MLE multilayered epithelium, IM intestinal metaplasia (Barrett’s esophagus), LGD low-grade dysplasia, HGD high-grade dysplasia

aNeeds confirmation by esophageal function tests (manometry, reflux monitoring), as described in the text

bCLE with increased cancer risk (IM, LGD, HGD) is treated (endoscopic mucosal resection ± radiofrequency ablation)

In advanced GERD cases, reflux-induced inflammation may also affect the distensibility of the esophagus, i.e., cause a functional outflow obstruction, which in turn causes CLE associated dysphagia. Therefore, after the exclusion of tumor induced dysphagia, impedance planimetry (Endoflip®) is recommended to objectify the amount of impaired distensibility as the cause for the functional outflow obstruction in the distal esophagus [89]. Endoflip operates an impedance sensing electrode surrounded by a balloon filled with an electrolyte solution. The diameter of the fluid correlates with the impedance. During the test, the balloon is placed along the esophagogastric junction. Upon fluid infusion the balloon circumferentially extends according to the distensibility of the esophagogastric junction [89, 90]. Consequently the planimetric resistance reflects the distensibility for a given level. Computer generated color plots remodel the distensibility contour along the esophagogastric junction in 1.0 cm increments. Thus cylindrical and hourglass contours of the planimetry color plot indicate normal and impaired distensibility of the esophagogastric junction, respectively. Indications for impedance planimetry include CLE with dysphagia (tumor excluded) and achalasia [89]. In addition the technology serves for therapy monitoring during and after functional surgery (antireflux surgery for CLE and GERD; dilatation or myotomy for achalasia) [89, 90]. Since Endoflip works independent from the electrophysiological quality of the esophageal surface (mucosal type) it can be used irrespective of the presence or absence of CLE.

Radiology

Radiologic tests for the assessment of CLE—related morphologies include videofluoroscopy [91], computed tomography ± positron emission tomography (PET) scan [92, 93] and magnetic resonance tomography imaging (MRI) ± contrast medium [94]. Esophageal endosonography is inferior to PET CT scan for the assessment of local tumor status and lymph node invasion [95]. Videofluoroscopy remodels the contour of the esophagogastric junction (suspicious for hernia) and indicates the presence of functional disorders (achalasia, esophageal spasm), diverticula, webs, rings, stenosis, polyps and tumors. Radiologic tests cannot assess CLE per se, but detect the consequences of CLE, including ulcer, tumor, and ring formation [91]. CT (± PET scan) and MRT assess intra and extra-esophageal manifestations of CLE related pathologies including the detection of the primary tumor (esophageal adenocarcinoma), local tumor infiltration, local and distant lymph node involvement, and metastasis formation (liver, lung). Thus imaging serves for tumor staging and treatment monitoring (assessment of response) [92, 93]. Radiologic tests do not replace endoscopy, biopsy sampling and histopathology of CLE. Furthermore, suspected functional diseases of the esophagus are to be confirmed by high resolution manometry. The application of MRI for the assessment of esophageal motility disorders is currently under investigation [94].

Discussion

Our findings indicate, that CLE results from life style and eating behavior and represents the morphologic consequence of gastroesophageal reflux. A subset of persons with CLE develop Barrett’s esophagus, which associates with an increased cancer risk [21, 22]. The implications of our findings for the therapy of CLE remain to be questioned.

Conceptually, impairment of life quality/productivity and/or cancer risk define disease [57]. Thus, CLE without symptoms and increased cancer risk is abnormal, but does not require treatment (Table 1) [12]. Surveillance endoscopy within 5 years is recommended (0.07 % annual cancer risk for CM/OCM) [56]. CLE with life quality/productivity impairing symptoms, but without cancer risk (i.e., absence of Barrett’s esophagus) define disease and require treatment (life style, medical, surgical) (Table 1) [12]. Antireflux surgery should only be offered to those who fulfill the following criteria: at least partial response to proton pump inhibitor (PPI) therapy; manometry assesses impaired function of the esophagus, i.e., incompetence of the LES (achalasia excluded); positive pH test or impedance pH test (reflux monitoring); is not willing to continue medical therapy; is fit for the operation (absence of contra-indications, comorbidities) [46, 96, 97]. If a patient is not suitable for antireflux surgery, PPI therapy and life style measures should be continued [98]. Both total (Nissen) and subtotal posterior fundoplication (Toupet) are equally effective to eliminate GERD symptoms [99, 100]. Fundoplication is superior to PPI therapy for the elimination of GERD symptoms at the cost of a somehow increased frequency of gas bloat and dysphagia [101].

Presence of CLE with low- and high grade dysplasia (reconfirmed by an expert pathologist) should be offered radiofrequency ablation (RFA; HALO®, BarrX, GI Solution,Covidien, USA) [102106]. Briefly, RFA represents a novel, endoscopic therapy for the effective and durable elimination of CLE with low- and high grade dysplasia [102106]. The radiofrequency energy is delivered toward the CLE via a catheter mounted balloon electrode (HALO® 360) for circumferential ablation or an endoscope mounted electrode (HALO® 60 and 90) for focal ablation. As a consequence the Barrett’s mucosa, but not deeper layers of the esophageal wall are ablated [102106]. RFA works in more than 80 to 90 % of the cases after 1 HALO® 360 and 2–3 subsequent focal ablations (HALO® 60, 90) [102106]. LGD and HGD harbor a 5–10 fold increased cancer risk, when compared to those with nondysplastic Barrett’s esophagus [107]. RFA durably prevents cancer in those with dysplasia [102, 103]. Therefore RFA (± endoscopic resection of endoscopically visible lesions) represents the recommended therapy for LGD and HGD and early cancer (T1a) [102, 103, 108]. Discrepancy exists, regarding the management of Barrett’s esophagus without dysplasia. Fleischer et al. suggested, that due to its biological characteristics, nondysplastic Barrett’s esophagus (NDBE) should be considered as a neoplasm [107]. Furthermore, NDBE shares numerous diagnostic and prognostic insecurities [107]. For an individual with NDBE we do not know how long it exists, if and when it will progress to cancer (time from NDBE to cancer development 5–15 years), if an advanced stage of CLE (dysplasia, early cancer) has been missed during biopsy sampling (biopsy sampling error), or if the pathologist made the correct diagnosis (inter-observer variation) [21, 22]. Furthermore, NDBE shares genetic characteristics of dysplasia and cancer (mutations, abnormal expression of growth factors, and their receptors) [21, 22, 109, 110]. Finally, the cancer risk of NDBE increases to that of dysplasia in the presence of one or more of the following prognostic factors: male gender, family history of gastrointestinal cancer, history of GERD symptoms > 10 years, hiatal hernia > 2 cm, endoscopically visible CLE > 3 cm and esophagitis [111]. The data justify to consider RFA for those with NDBE and the above risk profile within controlled trials. Following RFA patients are kept on high-dose PPI therapy [98, 102108]. Recent studies demonstrated that effective antireflux surgery favors the regression of NDBE and LGD [112114]. Going in line with these observations, the combination of RFA and antireflux surgery has been shown to increase the yield for NDBE elimination [115, 116]. Future studies are to be designed to compare the impact of PPI therapy, RFA, and antireflux surgery on Barrett’s esophagus and if these strategies might contribute to cancer prevention.

Current policies regarding Barrett’s esophagus and cancer prevention focus on those with symptomatic Barrett’s esophagus [4, 10, 21, 22]. As a matter of fact, 40–60 % of those with cancers arising within the SOG and the DDE (i.e., cardia) do not report a prior history of GERD symptoms before the development of cancer induced dysphagia [20]. Screening and surveillance of symptomatic Barrett’s esophagus only detects less than 10 % of all esophageal adenocarcinomas [21, 22, 26, 66]. However, surveillance assessed the cancers at a less advanced stage with the option for curative treatment, when compared to those detected due to tumor induced dysphagia [21, 22, 81]. The majority of these cancers present at an advanced stage and palliation remains the only therapeutic option [117]. In contrast, cancers assessed during surveillance associate with significantly longer survival [81]. How these observations translate into cancer prevention remains to be questioned.

Screening of the right population, i.e., which contains the majority of those at risk, contributes to detect the candidates for cancer development. Surveillance of those at risk offers the opportunity to eliminate the disease before it turns into cancer. Going in line with the above considerations we should extend screening to the population without GERD symptoms (the majority of cancers develop within this segment of the population) [20]. For both genders, screening endoscopy between 40 and 50 years of age harbors the highest yield for the assessment of NDBE [118]. CLE without NDBE (i.e., CM, OCM) should undergo surveillance in 5 years intervals (0.07 % annual cancer risk) [56]. Those with NDBE (0.12–0.7 % annual cancer risk) and the above listed risk profile should be offered RFA within clinical trials [111]. In absence of the risk profile we should offer surveillance at 3 years intervals [21, 22] and eliminate CLE at the stage of dysplasia, using RFA (± endoscopic resection). Cardiac and oxyntocardiac mucosa per se do not require treatment, unless these mucosal types are associated with reflux-induced impairment of the life quality. Here treatment is directed against the symptoms and aims to eliminate the complaints, either by life style measures [2, 4], medical therapy [98], or antireflux surgery [119].

Future studies will have to examine the impact of the above strategies for cancer prevention. Finally, the clinical relevance of novel molecular biology tools for the diagnosis and screening of Barrett’s esophagus is awaited (i.e., cytosponge, genetic markers [120, 121]). Remains to be awaited in as much novel endoscopic technologies with the opportunity to take optical biopsies will contribute to replace standard histopathology in the diagnosis of CLE [122].

Conflict of interest

The authors declare that there exists no conflict of interest.

Copyright information

© Springer-Verlag Wien 2013